Research Article CompaRob: The Shopping Cart Assistance Robotdownloads.hindawi.com/journals/ijdsn/2016/4781280.pdf · 2016. 2. 7. · e study carried out in [ ] concludes that elder

Research ArticleCompaRob The Shopping Cart Assistance Robot

Jorge Sales Jose V Martiacute Rauacutel Mariacuten Enric Cervera and Pedro J Sanz

Computer Science and Engineering Department Jaume I University Avenida Vicente Sos Baynat sn 12071 Castello de la Plana Spain

Correspondence should be addressed to Jorge Sales salesjujies

Received 2 August 2015 Revised 27 December 2015 Accepted 3 January 2016

Academic Editor Joo-Ho Lee

Copyright copy 2016 Jorge Sales et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Technology has recently been developed which offers an excellent opportunity to design systems with the ability to help peoplein their own houses In particular assisting elderly people in their environments is something that can significantly improve theirquality of life However helping elderly people outside their usual environment is also necessary to help them to carry out dailytasks like shopping In this paper we present a person-following shopping cart assistance robot capable of helping elderly peopleto carry products in a supermarket First of all the paper presents a survey of related systems that perform this task using differentapproaches such as attachable modules and computer vision After that the paper describes in detail the proposed system and itsmain featuresThe cart uses ultrasonic sensors and radio signals to provide a simple and effective person localization and followingmethod Moreover the cart can be connected to a portable device like a smartphone or tablet thus providing ease of use to the enduser The prototype has been tested in a grocery store while simulations have been done to analyse its scalability in larger spaceswhere multiple robots could coexist

1 Introduction

Recently the proliferation of big supermarkets and shoppingcenters added to the rapid development of robot technologyhas produced robotic systems for helping people in theseparticular environments Some specific types of applicationssuch as enhancing the physical capabilities of the user helpingin the transport of products by providing a mobile shoppingbasket and improving the information given to the user ina more intelligent manner have been described in scientificliterature

Assisting elder people in their environments is a way ofsignificantly improving their quality of life As an examplenetworks of sensors and actuators embedded in buildings(ie Ambient Intelligence) can provide useful functions Infact several European Union projects have focused on thisparticular subject However helping elderly people outsidetheir usual environment is also necessary to help them tocarry out daily tasks like shopping

Population ageing is a concern for most developed coun-tries and it is forecast that it will accelerate in the years tocome The main goal in the field of assistance robotics is tohelp users in their everyday lives in tasks such as moving

around handling objects or interacting possibly remotelywith other people either friends or relatives or professionalsIn this context robot devices are aimed not only at assistingpeople instead of replacing them but also at adopting the roleof companions Therefore they must be designed to interactwith untrained people and a great effort must be made toenable users to accept them easily

Moreover these assistance robots may have to operate inunstructured environments that cannot be customized suchas homes shops or parks This implies a need for the robotthat is conscious of its surroundings which requires the useof a variety of sensors In addition many of the things thatneed to be handled such as food do not come in standardsizes and shapes as they often do in an industrial context

In this paper we present a person following shoppingcart and assistance robot named CompaRob the companionrobot (Figure 1) capable of easily helping elderly peopleto carry products in a supermarket First of all the paperpresents a survey of related systems designed to help people insupermarkets and commercial malls [1ndash3] and to help peopleshopping and carrying their articles in a supermarket [4 5]using different approaches such as attachable modules andcomputer vision After that the paper describes in detail

Hindawi Publishing CorporationInternational Journal of Distributed Sensor NetworksVolume 2016 Article ID 4781280 15 pageshttpdxdoiorg10115520164781280

2 International Journal of Distributed Sensor Networks

Figure 1 CompaRob a person following shopping cart assistancerobot based on the Erratic robotic mobile platform equipped with ashopping basket

the proposed system a mobile platform that implementsperson following behaviour that uses ultrasound and radiotransmissions technology to provide robustness and ease ofuse from the point of view of the user Finally the paperanalyzes the scalability of this system for increasing thenumber of robots in a simulated environment and its usabilityin a small grocery store

2 Review of Developed Systems

21 Information and Help Systems In the context of infor-mation and help systems the services that those systemsprovide to the customers in malls and big supermarkets aremainly centered on indicating the location of certain shopsand offering nonspecific advertisement information based oncustomer identification

In [1] a system is described that has the ability toaccompany the user in the search of a list of articles previouslyspecified The system does not carry the food Moreover therobot performs navigation and obstacle detection by using acamera and a laser range finder

Using a Robovie IIF platform with humanoid aspect in[2] a system is described that interacts with the persons in amall and is able to give them directions about the localizationof the products or even recognizing them through RFID tagsThe user input is provided by means of a speech recognitionsystem

An intelligent multirobot system for big stores is pro-posed in [3] where the robots may develop cooperative tasksin order to detect available space in the shelves checking thereal location of the products by looking at their barcodesknowing the number of articles in a shelf by using RFID tagsmaking an inventory of the available products and helpingcustomers to localize specific articles

22 Shopping Cart Robots In recent years some studies haveaddressed the idea of developing an autonomous shoppingcart As an example the possibility of developing a porting

system for large-sized facilities like a shopping mall oran airport is studied in [4 5] To achieve this purposethey have attached a drive system to a shopping cart inorder to transmit power to the cart wheels To performthe desired person following capability they use a stereocamera installed in front of the cart One advantage of thissystem is that the person does not have to wear any attachedelement In contrast as they are using recognition by coloursegmentation the person has to wear clothes of a specificcolour Additionally they have to face all the problems relatedto computer vision high computational cost light variabilityproblems background colour interference problems and soforth

Amore complete prototype is proposed in [6] composedof two robots a companion and a carrier In order to localizethe position of the companion a set of cameras is used whichallows the detection of robots and humans present in thescenario The carrier robot is able to follow the companionone which is able to guide the customer to the desiredposition as well as follow him while waiting for the nextaction to be performed A system for tracking shopping cartsin indoor environments is developed in [7] This systemconsists of the installation of a computer and a video cameraon the shopping carts for them to be able to self-localizeand send their position to a centralized system We find thissystem interesting as it could also be a first step to having anautonomous cart with person following capabilities but as adisadvantage it requires a complex system to be installed onthe cart and a building with preexistent infrastructure

The study carried out in [8] concludes that elder peopleinteract in a better way with robots carrying the shoppingbasket when they have a humanoid aspect and provideconversational capabilities although this fact implies a highcost and complex system In [9 10] a shopping help systemis developed able to obtain the shopping list from a mobiledevice through a QR code carry the shopping basket andshow at each moment which articles are on it and communi-cate with the supermarket computer system to inform aboutthe location of articles It uses a laser range finder sonar andcontact sensors (bumpers) to navigate

A special case is presented in the system proposed in [11]where the user controls the robot remotely by specifying theproducts to be graspedThen the robot navigates through thesupermarket to perform the task autonomously A lot of workis provided in the design of the artificial hand used to grasp awide variety of food products No details are provided in thepaper about the localization and navigation techniques usedby the robot It seems a very useful system for handicappedpeople On the other hand some elderly people prefer beingpresent in the supermarket to touch the products and make amore detailed selection

23 Robot Person following Behaviour Person following per-formed by a robot (see Figure 2) has traditionally been animportant research topic in the machine vision area It isnatural to imagine that mobile robots of the future especiallythose operating in public places will be expected to have thisskill As an example in [12 13] two different efficient person-tracking algorithms for a vision-based mobile robot using

International Journal of Distributed Sensor Networks 3

Figure 2 Robot following an elder person using ultrasound andradio sensors An ultrasound ring transmitter is attached to the elderpersonrsquos legThe robot can follow the elder person using the receivedsignals

two independently moving cameras are presented With thisapproach it is possible to carry out real-time person followingin indoor environments In [14] the authors describe arobot with following function and returning function usinga monocular camera

A different human tracking method using templatematching in range data obtained from a laser range finder(LRF) is described in [15] The block matching is performedusing stored templates which correspond to appearancesof human legs In [16] the authors use laser scans andfiltering techniques for tracking moving objects and personsIn [17] the authors use a laser-based person-trackingmethodand two different approaches to person following direction-following and path-following and a combination of bothmethods is proposed a hybrid approach with the robotautomatically selectingwhichmethod to use In [18] the laser-based leg detector is combined with a Kinect system to detectthe user position predict its trajectory and make the vehiclefollow it In [19] the laser range finder is combined withtwo cameras to detect the user position and the presenceof obstacles while the main goal of the paper is to providethe robot with different behaviour strategies (eg Go ToUser Follow User Local Search and Global Search) andmotion ways (Free Space and Cluttered Space) selecting theappropriate one in every situation

A robot tracking system based on a camera and a light-emitting device is used in [20]Their purpose is to develop anautonomous mobile robot which can follow a human beingThe robot with a camera looks at a human having a light-emitting device The device consists of two LEDs The robotcan estimate the position of the human from the distancebetween two LEDs and the direction from the position ofthe LED on the captured image

In [21] a double sensor system is used to perform personfollowing by a robot The robot can accompany a personusing vision-based target detection and avoid obstacles withultrasonic sensors while following the person In [22] acamera is used to detect both the person to be followed and

the environmental obstacles while RFID tags are used toobtain the robotrsquos localization

In [23] a shopping help robot is proposed for visuallyimpaired people using RFID tags for product identificationand a laser range finder for localization and navigation whilein [24] RFID is mixed with visio techniques to identify theuser in a crowd and track his or her position in order to makeit possible to follow him

As can be seen several approaches have been developedusing different kinds of sensor information computer visionlaser range finder devices light-emitting diodes used as a lightpattern ultrasonic sensors to detect obstacles and so forth InTable 1 a comparison between the reviewed systems is shown

The system proposed in this paper is a shopping cartrobot that implements a person following behaviour Theobjective is to have a simple and robust system easy to usespecially for older people and that represents a relatively lowcost solution compared to other approaches The proposedperson following system is based on a combination of radioand ultrasound signals to perform measurements betweenthe person and the robot Once the system takes a pair ofmeasurements a simple operation estimates the position ofthe human to be followed The implementation requires acouple of measurement devices installed on the robot and anultrasound ring attached to the personrsquos leg This detectiontechnique as different to other implementations avoidsthe use of cameras and the large amount of computationnecessary to image processing

This systemhas already proven its efficiency in the contextof the EU GUARDIANS project [25 26] where the riskyand low visibility environments required a robust and reliablesystem Its strong points are the simplicity robustness and itsindependence to changes in visibility conditions Comparedto other proposed methods computational requirements arelow so it is possible to implement the whole system in a cost-effective programmable board

3 System Description The Shopping CartAssistance Robot

In this section the proposed shopping cart assistance robotnamed CompaRob is presented The temporal developmentof the system and the achieved milestones are representedin Figure 4 To the best of the authorsrsquo knowledge it is thefirst shopping assistance robot that combines modularitysimplicity and ease of use at a reasonable cost that is especiallytargeted to help the elderly people in such a common activitylike doing shopping As it will be described below it is basedon a commercial robotic platform and several accessorieslike an RFID tag reader that provides additional control ofthe products that are being added to the cart Moreoverwhen the system is connected to a mobile device like a tabletcomputer or smartphone it provides additional informationto the user and allows the adjustment of different personfollowing parameters

31 The Mobile Platform The CompaRob is based on theErratic-Videre mobile robot platform [27] a platform whichhas been equipped with a shopping basket It is a compact

4 International Journal of Distributed Sensor NetworksTa

ble1Com

paris

onbetweenmainfeatures

ofseveralreviewed

syste

ms

Sensors

Capabilities

Interaction

Camera

LRFlowast

RFID

Other

Follo

wing

Carryitems

Find

items

User

Environm

ent

Chen

andBirchfi

eld2007

[13]

Bino

cular

NO

NO

NO

NO

NO

NO

A

Chun

getal2012[15]

NO

NO

NO

NO

NO

NO

A

Cosgunetal2013[18]

NO

Kinect

na

NO

RemoteS

kype

chat

AFH

Garcia-Arroyoetal2012[9]

na

NO

BumpersQ

Rreaderson

arNO

NO

Disp

laykeyboard

FG

Germae

tal2009

[24]

Userd

etectio

nna

NO

NO

Touchpanel

F

GranataandBidaud

2012

[19]

Twocameras

NO

na

na

na

na

A

Grossetal200

9[1]

Omnidirectional

plus

frontal

NO

Sonarsensors

bumpers

NO

NO

Guide

user

totheitems

Touchpanel

speechhead

movem

ents

F

Kand

aetal2010

[2]

Bino

cular

NO

Person

identifi

catio

n

Speech

recogn

ition

floo

rsensors

NO

NO

Give

directions

Speecharm

and

head

movem

ents

H

Katic

andRo

dic2

010[3]

Inventorykeeping

Infrared

Hum

anassistance

AC

EFI

Kulyuk

inetal2005[23]

NO

Lo

calization

na

NO

Keypadspeech

D

Matsuhira

etal2010[6]

Stereo

plus

omnidirectional

NO

Touchandforce

sensorssonar

To

uchpanel

AFIJ

Nagum

oandOhya2

001[20]

NO

NO

UserL

ED

NO

NO

NO

A

Nish

imurae

tal2006

[4]

Stereo

Fu

ture

Encoder

NO

LEDs

AH

Shiehetal200

6[22]

Localization

na

na

LC

Dscreen

AD

Tomizaw

aetal2007

[11]

NO

na

NO

NO

Remotec

hat

FHK

Tsud

aetal2009

[14]

NO

NO

NO

Return

toorigin

NO

A

Yoshim

ietal2006

[21]

Bino

cular

NO

NO

Ultrason

ic

na

Touchpanel

speech

AF

Zend

eretal2007[16]

NO

NO

Odo

metry

na

NO

Speech

A

Zimmerman

2006

[7]

Mon

ochrom

atic

NO

NO

Barcod

efor

localizo

dom

Noautono

m

movem

ent

Not

implem

ented

Disp

layforitems

list

B

Com

paRo

b[35]

NO

Obstacle

avoidancelowastlowast

Identifyitems

UserR

Fsonar

NO

Smartpho

neta

blet

PCF

lowast

Laserrange

finder

lowastlowast

Optionalfeature

nainform

ationno

tavailable

Aobstacle

(and

user)d

etectio

nB

barcod

efor

localization

Cbarcod

efor

itemsidentificatio

nDR

FIDforlocalization

ERF

IDforitemsidentificatio

nFdataconn

ectio

nwith

warehou

se

Gdatac

onnectionwith

user

Htele

operation

Imutualinteractio

nwith

otherrob

ots

Jenvironm

entcam

eras

forlocalization

Kob

jectmanipulation


and powerful platform and is capable of carrying a full loadof robotics equipment including an integrated PC laserrange finder and other sensors The selection was motivatedbecause it is a simple platform reasonably priced highlyexpandable and robust Additionally this platform is veryflexible as it has been designed to work either with an on-board PC using the power supply from the robot batteries orwith a laptop with its own battery Additionally the hardwareis compatible with PlayerStage [28] and ROS (Robot Oper-ating System) [29] both open-source and widely employedin the robotics community The platform is equipped withthree 12V 7AH lead-acid batteries providing a reasonableautonomy for the described purpose (around 2 hours for anormal shopping use)The platformweighs 135 Kg includingthe base batteries and accessories The supported payload is20Kg

32 Ultrasound Localization System The localization systemused in the shopping cart robot has been adapted from theprevious results obtained in the field of person followingunder emergency situations and low visibility conditions[30 31] This system uses ultrasound-based signals to makemeasurements between two points (see Figure 3) It also usesradio signals for synchronization purposes The developedultrasonic localization system involves the following devices

(i) Hagisonic sonars (transmittersreceivers) two differ-ent models of ultrasonic sonar [32] have been studiedand used for implementing time difference of arrival(TDoA) in order to derive position informationThose sensors (HG-M40DAI and HG-M40DAII) areultrasonic object detectors and range finders thatoffer very-short-to-medium-range detectionThe AIImodel offers a special narrow directional response25sim30 degrees in vertical direction to minimize thereflection of unwanted sound waves

(ii) Radio modules standard radio transmitterreceiverworking at 433MHz those modules allow the modu-lation of digital signals up to 1 KHz at maximum dis-tances of around 20ndash30mThosewirelessmodules areintended to provide a synchronizationmechanism bysending real-time signals between the different robotsthat could be present

(iii) Handy Board general purpose board based on the68HC11 microcontroller [33] it is widely used in thefield of mobile robots for educational hobbyist andindustrial purposes

33 TDoAwithRF andUltrasound With this approximationwe try to determine the 2D position of a physical target (iea person) with respect to a mobile robot by using the timedifference of arrival of two different signals each one witha known propagation speed (ie radio and ultrasound seeFigure 5)

The first step was to determine the feasibility of the useof ultrasound signals to obtain distances In order to do thatseveral laboratory experiments have been done to evaluatesome available ultrasound transmittersreceivers using wires

P(x y)

d1 d2

dr

Figure 3 Position determination by trilateration The robot hasan estimation of the elderly position by doing two independentmeasurements

to connect both transmitter and receiverThose sensors (HG-M40DAI and HG-M40DAII) are ultrasonic object detectorsand range finders that offer very-short-to-medium-rangedetection The AII model offers a special narrow directionalresponse 25ndash30 degrees in vertical direction to minimizethe reflection of unwanted sound waves We obtain largertransmission distances by removing the frontal cover of thesensors that acts as an anisotropic filter

After performing those experimentations we concludedthat the measurement of the time of flight of ultrasoundwaves could be a feasible way of estimating distances Theestimation of distances by measuring the time of propagationof ultrasonic waves can be useful for several localizationmethods based on the knowledge of some distances Furtherexperimentation demonstrated that this technique offereda good performance of the sensors for distances up to 7-8meters

The next step was to introduce synchronization radio sig-nals in order to estimate distances between two independentnodes (ie a robot and the person or two robots) To achievethis purpose the theoretical idea of the time difference ofarrival (TDoA) was implemented (see Figure 5) The idea isto measure the time of flight of the ultrasound signals usingthe radio signal as synchronizationmethodThe time of flightof the radio signal is considered constant for the range ofdistances that we are going tomeasure (from 0 to amaximumof 8 meters)

The most challenging task for implementing the TDoAwas to find a hardware platform that could take those mea-surements being able to control both radio and ultrasoundsensors After exhaustive research about its capabilities theHandy Board [33] widely used for teaching purposes inJaume I University was used Some additional electronicswere necessary for coupling those components to the board(see Figure 6) The programming of the board was done witha mixture of C programming language for the global routineand assembly language for the real-time reading and writingof signals


dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer

Figure 4 CompaRob shopping cart temporal development evolu-tion and summary of the achieved milestones

t1

t2

d

Figure 5 TDoA principle for measuring distances between robotsby using the time difference of arrival of two different signals (elec-tromagnetic and ultrasonic) each one with a known propagationspeed

Measurement system boardUltrasound module

Rx radio module

Figure 6 Radioultrasound measurement system module imple-mented with a Handy Board a general purpose board based on the68HC11 microcontroller

The measurement capabilities of the boards were testedand the result is that the obtained measurements were quiteaccurate being able to measure up to 8 meters with amaximum error of around 3 cm

34 Trilateration for Position Determination The prototypepresented in this paper uses a couple of measurementboards to determine the 2D position of the person The tworadioultrasound measurement systems (see Figure 7) areseparated by a known distance 119889

119903

Measurementunits

Figure 7 Erratic robot prototype to be used as a shopping cartequipped with a couple of radioultrasound measurement systems

With this configuration and using an algebraic trilater-ation method (see [34] for a nonalgebraic approach basedon constructive geometric arguments) we can measure thedistances from an emitter located at 119875 point (see Figure 8)The119875 point is the position of the person that wewant to locate(ie the elderly see Figure 3)

In order to obtain the (119909 119910) coordinates of the 119875 pointwe proceed as follows the 119875 point is in the intersection ofthe two circumferences 119862

1and 119862

2 Let the equation of the

circumference be (119909 minus 119886)2 + (119910 minus 119887)2 = 1199032 where (119886 119887) isthe center 119903 is the radius and (119909 119910) are the points of thecircumference

Knowing both center points (minus1198891199032 0) and (119889

1199032 0) and

the distances 1198891 1198892from the object located at 119875(119909 119910) we can

write the following system of equations

(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)

Solving this system of equations we obtain the point119875(119909 119910) where the emitter is located

119909 =1198892

1minus 1198892

2

2119889119903

119910 =

radicminus11988941+ 211988921(11988922+ 1198892119903) minus 11988942+ 1198892119903(211988922minus 1198892119903)

2119889119903

(2)

The equation for the 119910 coordinate provides two solutions(positive and negative) but we only use the positive solutionbecause ultrasound sensors are only receiving signals fromboth 119910 positive quadrants as they are mounted in positive 119910direction and not in negative 119910 direction Once we know therelative position of the 119875 point with respect to the robot acontrol algorithm computes the resultant linear and angularspeeds for controlling the robot


P(x y)

C1

C2

d1 d2

(0 0)(minusdr2 0) (dr

2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)

Figure 8 Trilateration principle (a)The 119875 point in the intersection of 1198621and 119862

2 (b) Uncertainty area due to measurements errors of 119889

1and

1198892

35 Trilateration Geometrical Error Study In order to assessthe error produced by individual measurements when usingthe above trilaterationmethod a geometrical study wasmadefor the most common configurations

When using the implemented trilateration techniquetwo independent measurements 119889

1and 119889

2are taken by

two measuring devices (see Figure 8(a)) from the receivingpoints (minus119889

1199032 0) and (119889

1199032 0) to the object located at119875(119909 119910)

equipped with a transmission device Individual measure-ment errors in 119889

1and 119889

2produce an uncertainty area (see

Figure 8(b)) when calculating 119875(119909 119910) by trilaterationAs an example in Figure 9 we present a geometrical

representation of the uncertainty areas in which the point119875(119909 119910) could be if we consider an individual error foreach measurement of plusmn5 cm (an upper bound typical erroris around plusmn1 cm) The uncertainty area is represented forincreasing distances from the emitter to the receivers (from1 to 8 meters) In this figure the emitter is situated at 40∘from the perpendicular to the segment that connects thetwo receivers When increasing the distance the uncertaintyarea becomes bigger for the horizontal coordinate (becomeswider) but remains small for the vertical coordinate Thehorizontal uncertainty is asymmetric due to the angulardisplacement of the emitter The error region increases withthe distance from the transmitter to the receivers but it can bereduced by increasing the separation distance (119889

119903) between

receivers that in practice is limited by the robot dimensionsEven this theoretical error study reveals that the position

determined by the method can be affected by an impor-tant horizontal error in practice the individual measuresobtained from the devices are much smaller than the theoret-icalmaximum so the position of the person can be accuratelydetermined and tracked by using a simple filtering technique

36 Attachable Ultrasound Ring The shopping cart assis-tance robot is able to follow an elderly person that has anultrasound ring attached to his or her leg (see Figure 10)This prototype device is composed of a 9V battery a radiotransmitter and five ultrasound transmitters The device is

800

700

600

500

400

300

200

100

014 ∘

17 ∘

EmitterUncertainty areaRobot with receivers

Figure 9 Trilateration geometrical error study when the emitteris moving away in diagonal line Distance between receivers50 cm Emitter-Robot distance 1ndash8m Emitter Angle 40∘ Angularuncertainty 14 + 17∘

continuously emitting radio and ultrasound signals in orderto indicate to the assistance robot the position of the elderlyperson

The particular design of the prototype was chosen sothat it can be easily attached to the leg and simple in usagefrom the point of view of the user (only requires to usean onoff switch) Its design is simple and discrete and itssimplicity implies also reliability It is also important to takeuser acceptability into account especially when designingproducts that are designed for the elderly


Figure 10 An ultrasound ring transmitter is attached to the elderlypersonrsquos leg The robot can follow the person using the receivedsignals

After testing the system with a small pilot group andgathering the opinion of several potential users we canconclude that the device is well accepted among elderlypeople It would even be desirable not having to use anyattached device the ultrasound ring is a necessary elementfor the robot to work and it does not affect in any way themobility of the person

37 RFID Tag Reader The CompaRob has been equippedwith an RFID tag reader allowing it to keep control of theproducts that are being added to the basket (see Figure 11)The list of products is sent to a user interface (described inSection 38) so the user can see the list of purchased productsand their total cost The tag reader is a Skyetek SkyeModuleM2 which connects to the on-board computer using a USBinterface

The software architecture includes a C++ server runningin the robotrsquos on-board computer which provides a TCPIPinterface to an Android device At the moment of writingthe server is running on a Windows Virtual Machine dueto some drivers issues which made it impossible to runthe server directly on the Linux operating system and ROSplatform

38 User Interface The CompaRob is connected wirelesslyto a tablet computer or smartphone allowing the user tocheck the shopping list Moreover the user can control someparameters of the robot related to the person followingfunctionality (described in Section 4) by means of a setupconfiguration screen

In this context the user is able to for example adjustthe following distance as well as modifying the followingspeed and the robot turn delay (see Figure 12)This will allowthe user to tune the way heshe wants the robot to performMoreover by using this interface the user can know otherrobot parameters like the battery status and so forth

4 Person following Functionality

The person following functionality is implemented in acomputer program allocated on the on-board PCof the robot

The PC runs an Ubuntu 1204 LTS distribution and the robotis controlled using ROS thanks to the erratic player node thatwraps the Erratic driver found in Player project The systemfunctionality is partially implemented in Java (the interface tothemeasurement units and the positioner) andPython (usingrospy a pure Python client library for ROS) to implement theperson following functionality and to interface with ROS andthe hardware platform controller

The person following functionality is implemented usinga 4-state machine (see Figure 13) Each state represents onedifferent position of the followed person with respect to therobot (see Figure 14) The 4 states are described as follows

(i) Follow the current position 119875(119909 119910) of the followedperson is into the working area tagged as Follow (seeFigure 14) The control program computes the robotspeed in a way that the position of the person 119875(119909 119910)converges to the desired position119875(1199091015840 1199101015840) To achievethis goal the rotation speeds of both robot wheels arecomputed using a proportional controller

(ii) Wait 1 the current position 119875(119909 119910) of the followedperson is into the nearest area tagged as Wait 1 at adistance from the robot that is less than that desiredThis situation happens for example when the personcomes back to take something else from the shelvesor when the person puts something into the cartIn this state the robot can behave in two differentways depending on the user configuration (1) to stopmoving so that the person can approach the robot toput something on the basket or (2) to slightly movebackwards providingmore freedom to the user whenmoving forwards and backwards In this latter casethe user can leave things in the basket by approachingto the robot from the side (Wait 2 state)

(iii) Wait 2 in this state the person has moved to the sideof the robot and leaves the coverage area of ultrasoundreceivers This situation happens for example whenthe person wants to leave something on the basketor when the person changes direction In this casethe robot waits for a configurable period of time(turn delay) just in case the person comes backto the coverage area If the person comes back tothe coverage area the new state will be Wait 1 orFollow Otherwise after waiting an amount of timethe program will transition to the Search state

(iv) Search the robot starts rotating slowly in order tosearch for the person This situation happens forexample after the personhas changed directionOncethe person has been located the possible next stateswill be Follow orWait 1

When the program starts its execution both measure-ment units installed on the robot are queried Depending onthe read value the initial state will be one of the followingFollowWait 1 orWait 2 After the initialization the programwill evolve as described above


(a) (b)

Figure 11 RFID tag reader allowing keeping control of the products that are being added to the basket (a) The list of products is sent to atablet computer (b)

(a) (b)

Figure 12 User interface to get the actual productsrsquo list (a) and adjusting some parameters related to the robot person following navigation(b)

Follow Wait_1

Wait_2Search

Figure 13 Person following behaviour 4-state machine Each staterepresents a different position of the followed person with respect tothe robot

5 Simulation Results

Several 2D simulations of the robots using PlayerStage andROShave been carried out to test the feasibility and scalability

of the presented system and are described below The videosof the simulations are available online [35]

51 Small Grocery Store Simulation The first simulationrecreates a small grocery store (see Figure 15) in which thereare five people doing shopping and five robots each onefollowing one of the five persons The environment is basedon a small grocery consisting of three corridors with shelveson each side The customers move along the corridors andstop in front of the shelves to pick a product

We were particularly interested in the scalability of oursystem that is how would the system perform as the numberof robots increases Due to the interferences between thesonar sensors of different robots the frequency of the pulses isdecreased In this simulation the positioning system providesnew measurements at a rate of merely 2Hz Such lowfrequency allows the group of robots to make measurementsof each person in turn without overlapping

Robots are programmed to follow the persons at a targetdistance of 1m Each robot follows the person depicted inthe same colour Figure 16 presents the distances between


Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)

Figure 14 Person following behaviour the person can be in one ofthe 3 shown regions corresponding to Wait 1 Wait 2 and Followstates

Figure 15 Simulation of five persons and robots on a small groceryconsisting of three corridors with shelves on each side

each person and the corresponding follower robot during aninterval of 70 seconds In such an interval a common personperforms between three and five stops and motions

When a person starts to move the distance to the robotincreases momentarily with peaks at about 22ndash24m sincethe velocities of robots are limited and the feedback controlloop is purely proportional But the positioning systemguidesthe robot towards the person following the appropriate pathuntil the person stops again in front of a shelf

Upon the person approaching the robot below a giventhreshold it will move slowly backwards to keep the prese-lected distance (eg see the distance of the robot r 5 in theplot between seconds 60 and 70) according to the personfollowing functionality described in Section 4

Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)

Figure 16 Person-robot distances between each person and thecorresponding follower robot during an interval of 70 seconds

52 Large Grocery Store Simulation The previous followeralgorithm works only for small distances We have testeda more complex follower behaviour in a larger shoppingenvironment Figure 17 depicts the new simulation environ-ment consisting of a larger shop with a higher numberof corridors and shelves The robots are programmed tocalculate a trajectory towards the position of the person giventhe map of the environment and a proper self-positioningsystem We have run a simulation of the five people doingshopping over a period of six minutes Distances betweeneach person and the corresponding follower robot are shownin Figure 18 Each robot updates the position of the followedperson at 1Hz The preprogrammed distance is 1m and mostof the time the robots remain within a distance to the personshorter than 25m But the distance may increase due to thedelay in the update and the moving obstacles (other personsandor robots) With our available sensors the positioningof the person can be reliably determined up to a distance of8m (threshold shown in the figure) Two robots (r 1 and r 4)approach such distance in two instants of the simulation Insuch cases the robot should stop and wait for the person tocome within the detection distance once more

6 Field Experiments (I) Person followingFunctionality in a Small Grocery Store

To evaluate the usability feasibility and most convenientrobot following parameters in a real environment severalexperiments were carried out in a small grocery store(Figure 19) The system was tested with a small pilot groupof elderly people with the aim to get some feedback relatedto usability and adequate robot parameters All of them wererequired to do their usual shopping following approximatelythe trajectory shown in Figure 20

Two different linear following speeds were selected firstat 02ms and then at 03ms following the preferencesof the users For each selected speed the distance fromthe robot to the person was recorded during an intervalof 70 seconds as in the first simulation experiment The


Figure 17 Simulation of five people shopping in a larger environ-ment

r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)

Figure 18 Person-robot distances in the large shop environment Adetection threshold for the ultrasound sensor is also shown

Figure 19The shopper with the CompaRob in a small grocery storeused as a real validation scenario

aim now is to test the robustness of the proposed followingmethod at low measuring rates (ie 2Hz) coming fromthe positioning system Also different following distanceswere tested 119889 = 02 05 10 to observe the behaviour atdifferent linear speeds As can be seen in Figure 21 (linearspeed 02ms) when the person starts to move the distanceto the robot increases and the robot tries tominimize it (using

Figure 20 Approximate floor plan of the grocery store the realvalidation scenario The trajectory followed by the shopper and therobot is marked in a dotted line

d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)

Figure 21 Person-robot distances in the real grocery environmentfor different following distances 119889 = 02 05 10 while a persondoes shopping The linear following speed was set to 02ms

a simple proportional controller) to the preselected followingdistances depicted in horizontal lines The small ripplepattern that can be observed in all the graphs correspondsto the walking pattern as the ultrasonic ring was attachedto the personrsquos leg This pattern does not affect the followerfunctionality in any way

As can also be observed when the person does a stopto pick something up the distance decreases as the robotapproaches the person This happens faster in Figure 22 asthe linear speed is 03ms slightly greater than in the exper-iments depicted in Figure 21 (02ms) When the personapproaches the robot for example to put something intothe basket the robot remains static if the user approachesfrom the side according to the state machine explained inSection 4 Contrarily if the user approaches from the frontthe robot begins to move slowly backwards to keep thepredefined distance providing more freedom to the userwhen moving along the corridor

For the tested following distances the robot behavesadequately but if greater distances were required (ie a userthat prefers a greater robot distance) an obstacle avoidancesystem should be used to avoid collisions To tackle thisissue the platform is also equipped with a Hokuyo URG-04LX Scanning laser range finder (LRF) [36] and uses a


d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)


local planner for obstacle avoidance based on the DynamicWindow Approach [37] to local navigation on a planeGiven the reference position of the person to follow and acostmap obtained by a laser range finder the local plannerproduces velocity commands to send to the mobile base Theinformation for the sonar needs to be combined with thelaser measurements in order for the detected person not tobe treated as an obstacle Nevertheless with the use of an LRFthis device increases significantly the cost of the platform soas an alternative a much cheaper optical triangulation-baseddistance sensor (like the Sharp GP2D12 distance measuringsensor) could be used instead

Further details like a video with the person followingsystem in action and some preliminary usability results canalso be seen online [35]

7 Field Experiments (II) Usability Tests withTwo Robots in a Large Grocery Store

Based on the first user experiments and suggestions animproved version of the robot has been designed and a firstprototype (named CompaRob-2 see Figure 23) was tested ina larger grocery store (see Figure 24) In the evolved versiona standard shopping cart bag was fitted to the robot bymeans of a PVC structure improving the ergonomics and theaestheticsThe previous prototype had a 22 times 25 times 40 cm (119867times119871times119882) bag while the improved one is 65times 29times 44 cm (see thedetails in Figure 25)The upper part of the PVC structure canbe used as a handle both to slightly lean on it and to move thecart Also the height of the bag has been increased making iteasier to put articles inside and increasing as well the capacityof the cart Below the bag a specific protected space has beenenabled to hold the computer and sensors

In this second experiment the new platform was tele-operated using a remote laptop connected wirelessly to thecomputer on board the robot and using the ROS teleoppackage The objective at this prototyping stage was to get asmuch feedback as possible from users before the design and

44 cm 29 cm

65

cm

Figure 23 CompaRob-2 prototype an improved version of Com-paRob shopping cart based on first user usability tests and sugges-tions

Figure 24 Usability tests with two robots in a larger grocery storeBoth CompaRob andCompaRob-2 robots were tested with differentpeople doing normal shopping

development of an improved person following functionalityFive different people were asked to do normal shopping ina grocery store acting as natural as possible and forgettingabout having to push the shopping cart

The user experiences were collected in a series of inter-views All of them indicated that this new version is more


25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)

Figure 25 Working drawings and dimensions of the original CompaRob (a) and the new CompaRob-2 robot (b)

Appropriateness of the attachable leg ring

Appropriateness of the use of a hand watch

Overall satisfaction with the CompaRob robot

Overall satisfaction with the CompaRob-2 robot

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

Figure 26 Summary of the responses given by the four interviewedusers related to usability and overall satisfaction with the twoversions of the robot

comfortable than the previous one mainly with the newdesign of the bag A summary of the main ideas brought bythe users in order to improve the system is the following

(i) Make the robot even more light in order to allow theuser to move it manually in case of necessity

(ii) Design a tool or elevator that helps elder people toextract the goods in a simple and easy manner

(iii) The use of an intelligent arm watch instead of anattachable belt to the leg

Moreover a set of four questions was asked to the fourpeople that tested the robots Their responses to a Likert1ndash5 scale are summarized in a chart (Figure 26) As can beseen all of them are very satisfied with the new design andthey mainly suggest the use of an intelligent arm watch as aninteraction device with the robot

8 Conclusion and Future Work

In this paper we have presented a person following systemfor a mobile assistance robot that is based on a combinationof radio and ultrasound signals to perform measurementsbetween the person and the robot One of the objectives ofthe work was to develop a simple reliable and easy to usesystem that could provide freedom of movements for theelderly people To do so the mobile robot follows the personby reading the signals from an ultrasound ring attached to thepersonrsquos leg

A second important objective was the user acceptabilityAfter testing the system with some potential users we wereable to conclude that this shopping cart robot is well acceptedamong elderly people because they understand that it is animprovement in their lives a technological advance insteadof something that replaces their physical capabilities

As future work we are planning to integrate the recentlyproposed indoor radio-based localization system [38] intothe shopping cart This system is based on fingerprintingtechniques and can be applied to robot and person local-ization in order to know their position in a particular envi-ronment [39] just using simply the WiFiZigBee transceiverinformation and providing a 15m localization error Thiscould improve the robot navigation inside a shopping centerand the robot could also show the person where to go inorder to find a particular product Combining the radiosonarlocalization for navigation and the fingerprinting globalapproach more sophisticated applications will be possible Infact further work can for example provide a more intuitiveuser interface by providing the possibility of searching for theavailable products in the store creating a wish list and let therobot conduct the user to the proper place in an efficient andautonomous manner This would require a better integrationof the robot interface with the store database On the otherhand if the user wants to lead the navigation the smartphone


can provide orientations to find the products related to theactual position of the robot and the user

Moreover some experiments are being performed toenhance the global localization system of the person by usingthe eZ430-Chronos watch [40] a device that enables the inte-gration of a sophisticated radio-based localization system andprovides a bidirectional communication link between userand the robot

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

This work was partly supported by Spanish Ministry underGrant DPI2014-57746-C3 (MERBOTS Project) and by Uni-versitat Jaume I Grants P1-1B2015-68 and PID2010-12

References

[1] H-M Gross H Boehme C Schroeter et al ldquoTOOMASinteractive shopping guide robots in everyday usemdashfinal imple-mentation and experiences from long-term field trialsrdquo in Pro-ceedings of the IEEERSJ International Conference on IntelligentRobots and Systems (IROS rsquo09) pp 2005ndash2012 IEEE St LouisMo USA October 2009

[2] T Kanda M Shiomi Z Miyashita H Ishiguro and N HagitaldquoA communication robot in a shoppingmallrdquo IEEETransactionson Robotics vol 26 no 5 pp 897ndash913 2010

[3] D Katic and A Rodic ldquoIntelligent multi robot systems forcontemporary shopping mallsrdquo in Proceedings of the 8th Inter-national Symposium on Intelligent Systems and Informatics (SISYrsquo10) pp 109ndash113 Subotica Serbia September 2010

[4] S Nishimura K Itou T Kikuchi H Takemura and HMizoguchi ldquoA study of robotizing daily items for an auto-nomous carrying systemmdashdevelopment of person followingshopping cart robotrdquo inProceedings of the 9th International Con-ference on Control Automation Robotics and Vision (ICARCVrsquo06) pp 1ndash6 Singapore December 2006

[5] S Nishimura H Takemura and H Mizoguchi ldquoDevelopmentof attachable modules for robotizing daily items -person fol-lowing shopping cart robot-rdquo in Proceedings of the IEEE Inter-national Conference on Robotics and Biomimetics (ROBIO rsquo07)pp 1506ndash1511 IEEE Sanya China December 2007

[6] NMatsuhira FOzaki S Tokura et al ldquoDevelopment of robotictransportation systemmdashshopping support system collaboratingwith environmental cameras and mobile robotsrdquo in Proceedingsof the 41st International Symposium on Robotics (ISR rsquo10) and6th German Conference on Robotics (ROBOTIK rsquo10) pp 1ndash6Munich Germany 2010

[7] T G Zimmerman ldquoTracking shopping carts using mobilecameras viewing ceiling-mounted retro-reflective bar codesrdquo inProceedings of the IEEE International Conference on ComputerVision Systems (ICVS rsquo06) p 36 IEEE January 2006

[8] Y Iwamura M Shiomi T Kanda H Ishiguro and N HagitaldquoDo elderly people prefer a conversational humanoid as ashopping assistant partner in supermarketsrdquo in Proceedings

of the 6th ACMIEEE International Conference on Human-Robot Interaction (HRI rsquo11) pp 449ndash457 Lausanne SwitzerlandMarch 2011

[9] M Garcia-Arroyo L F Marin-Urias A Marin-Hernandezand G D J Hoyos-Rivera ldquoDesign integration and test ofa shopping assistance robot systemrdquo in Proceedings of the 7thAnnual ACMIEEE International Conference on Human-RobotInteraction (HRI rsquo12) pp 135ndash136 ACM Boston Mass USAMarch 2012

[10] A Marin-Hernandez G de Jesus Hoyos-Rivera M Garcıa-Arroyo and L F Marin-Urias ldquoConception and implementa-tion of a supermarket shopping assistant systemrdquo in Proceedingsof the 11th Mexican International Conference on Artificial Intel-ligence (MICAI rsquo12) pp 26ndash31 IEEE San Luis Potosi MexicoNovember 2012

[11] T Tomizawa K Ohba A Ohya and S Yuta ldquoRemote foodshopping robot system in a supermarketmdashrealization of theshopping task from remote placesrdquo in Proceedings of theIEEE International Conference onMechatronics and Automation(ICMA rsquo07) pp 1771ndash1776 IEEE Harbin China August 2007

[12] H Kwon Y Yoon J B Park and A C Kak ldquoPerson trackingwith a mobile robot using two uncalibrated independentlymoving camerasrdquo in Proceedings of the IEEE InternationalConference on Robotics and Automation (ICRA rsquo05) pp 2877ndash2883 Barcelona Spain April 2005

[13] Z Chen and S T Birchfield ldquoPerson following with a mobilerobot using binocular feature-based trackingrdquo in Proceedingsof the IEEERSJ International Conference on Intelligent Robotsand Systems (IROS rsquo07) pp 815ndash820 San Diego Calif USANovember 2007

[14] N Tsuda S Harimoto T Saitoh and R Konishi ldquoMobile robotwith following function and autonomous return functionrdquo inProceedings of the ICROS-SICE International Joint Conference(ICCAS-SICE rsquo09) pp 635ndash640 Fukuoka Japan August 2009

[15] W Chung H Kim Y Yoo C-B Moon and J Park ldquoThe detec-tion and following of human legs through inductive approachesfor a mobile robot with a single laser range finderrdquo IEEE Trans-actions on Industrial Electronics vol 59 no 8 pp 3156ndash31662012

[16] H Zender P Jensfelt and G-J M Kruijff ldquoHuman- andsituation-aware people followingrdquo in Proceedings of the 16thIEEE International Conference on Robot and Human InteractiveCommunication (RO-MAN rsquo07) pp 1131ndash1136 JeJu Island SouthKorean August 2007

[17] R Gockley J Forlizzi and R Simmons ldquoNatural person-following behavior for social robotsrdquo in Proceedings of theACMIEEE Conference on Human-Robot Interaction (HRI rsquo07)pp 17ndash24 ACM Washington DC USA March 2007

[18] A Cosgun D A Florencio and H I Christensen ldquoAuto-nomous person following for telepresence robotsrdquo in Pro-ceedings of the IEEE International Conference on Robotics andAutomation (ICRA rsquo13) pp 4335ndash4342 IEEE Karlsruhe Ger-many May 2013

[19] CGranata andP Bidaud ldquoA framework for the design of personfollowing behaviors for social mobile robotsrdquo in Proceedingsof the 25th IEEERSJ International Conference on Robotics andIntelligent Systems (IROS rsquo12) pp 4652ndash4659 IEEE VilamouraPortugal October 2012

[20] Y Nagumo and A Ohya ldquoHuman following behavior ofan autonomous mobile robot using light-emitting devicerdquo inProceedings of the 10th IEEE International Workshop on Robot


and Human Interactive Communication (ROMAN rsquo01) pp 225ndash230 Bordeaux France September 2001

[21] T Yoshimi M Nishiyama T Sonoura et al ldquoDevelopment ofa person following robot with vision based target detectionrdquoin Proceedings of the IEEERSJ International Conference onIntelligent Robots and Systems (IROS rsquo06) pp 5286ndash5291 IEEEBeijing China October 2006

[22] M-Y Shieh Y-H Chan Z-X Lin and J-H Li ldquoDesign andimplementation of a vision-based shopping assistant robotrdquo inProceedings of the IEEE International Conference on SystemsMan and Cybernetics (SMC rsquo06) pp 4493ndash4498 IEEE TaipeiTaiwan October 2006

[23] V Kulyukin C Gharpure and J Nicholson ldquoRobocart towardrobot-assisted navigation of grocery stores by the visuallyimpairedrdquo in Proceedings of the IEEERSJ International Confer-ence on Intelligent Robots and Systems (IROS rsquo05) pp 2845ndash2850 Edmonton Canada August 2005

[24] T Germa F Lerasle N Ouadah V Cadenat and M DevyldquoVision and RFID-based person tracking in crowds from amobile robotrdquo in Proceedings of the IEEERSJ International Con-ference on Intelligent Robots and Systems (IROS rsquo09) pp 5591ndash5596 IEEE St Louis Mo USA October 2009

[25] GUARDIANS ldquoGroup of unmanned assistant robots deployedin aggregative navigation supported by scent detection(GUARDIANS)rdquo EU Project IST-045269 2007 httpswwwshuacukresearchmeriresearchguardians-project

[26] J Sales R Marn E Cervera S Rodrıguez and J Perez ldquoMulti-sensor person following in low-visibility scenariosrdquo Sensors vol10 no 12 pp 10953ndash10966 2010

[27] Erratic-Videre Erratic-Videre Mobile Robot Platform Erratic-Videre 2010 httpwikirosorgRobotsErratic

[28] Player ldquoThe Player Project Free Software tools for robot andsensor applicationsrdquo 2010 httpplayerstagesourceforgenet

[29] M Quigley K Conley B P Gerkey et al ldquoROS an open-sourceRobot Operating Systemrdquo in Proceedings of the ICRAWorkshopon Open Source Software Kobe Japan May 2009

[30] J Sales M El-Habbal R Marın et al ldquoLocalization of net-worked mobile sensors and actuators in low-visibility condi-tionsrdquo in RISE (IARPEURON Workshop on Robotics for RiskyInterventions and Environmental Surveillance) SheffieldHallamUniversity Sheffield UK 2010

[31] A Naghsh J Saez-Pons J Penders et al ldquoRemote and in-situ multirobot interaction for firefighters interventions undersmoke conditionsrdquo in Proceedings of the 2nd IFAC Symposiumon Telematics Applications (TA rsquo10) pp 109ndash115 TimisoaraRumania October 2010

[32] Hagisonic Ultrasonic Sensors Hagisonic 2010 httpwwwhagisoniccom

[33] Handy-BoardTheHandyBoard 2010 httpwwwhandyboardcom

[34] F Thomas and L Ros ldquoRevisiting trilateration for robot local-izationrdquo IEEE Transactions on Robotics vol 21 no 1 pp 93ndash1012005

[35] CompaRob ldquoCompaRob the shopping cart assistancerobotrdquo 2015 httpsitesgooglecomaujieselderlyassistancecomparob

[36] Hokuyo Hokuyo URG-04LX Scanning Laser Range Finder2010 httpwwwhokuyo-autjp02sensor07scannerurg 04lxhtml

[37] D Fox W Burgard and S Thrun ldquoThe dynamic windowapproach to collision avoidancerdquo IEEERobotics andAutomationMagazine vol 4 no 1 pp 23ndash33 1997

[38] J VMarti J Sales RMarin and E Jimenez-Ruiz ldquoLocalizationofmobile sensors and actuators for intervention in low-visibilityconditions the ZigBee fingerprinting approachrdquo InternationalJournal of Distributed Sensor Networks vol 2012 Article ID951213 10 pages 2012

[39] J V Martı J Sales R Marın and P J Sanz ldquoMulti-sensorlocalization and navigation for remote manipulation in smokyareasrdquo International Journal of Advanced Robotic Systems vol10 article 211 2013

[40] A Fink J Lange and H Beikirch ldquoRadio-based indoorlocalization using the eZ430-Chronos platformrdquo in Proceedingsof the IEEE 1st International Symposium on Wireless Systems(IDAACS-SWS rsquo12) pp 19ndash22 IEEE Offenburg GermanySeptember 2012

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



Figure 1 CompaRob a person following shopping cart assistancerobot based on the Erratic robotic mobile platform equipped with ashopping basket

the proposed system a mobile platform that implementsperson following behaviour that uses ultrasound and radiotransmissions technology to provide robustness and ease ofuse from the point of view of the user Finally the paperanalyzes the scalability of this system for increasing thenumber of robots in a simulated environment and its usabilityin a small grocery store

2 Review of Developed Systems

21 Information and Help Systems In the context of infor-mation and help systems the services that those systemsprovide to the customers in malls and big supermarkets aremainly centered on indicating the location of certain shopsand offering nonspecific advertisement information based oncustomer identification

In [1] a system is described that has the ability toaccompany the user in the search of a list of articles previouslyspecified The system does not carry the food Moreover therobot performs navigation and obstacle detection by using acamera and a laser range finder

Using a Robovie IIF platform with humanoid aspect in[2] a system is described that interacts with the persons in amall and is able to give them directions about the localizationof the products or even recognizing them through RFID tagsThe user input is provided by means of a speech recognitionsystem

An intelligent multirobot system for big stores is pro-posed in [3] where the robots may develop cooperative tasksin order to detect available space in the shelves checking thereal location of the products by looking at their barcodesknowing the number of articles in a shelf by using RFID tagsmaking an inventory of the available products and helpingcustomers to localize specific articles

22 Shopping Cart Robots In recent years some studies haveaddressed the idea of developing an autonomous shoppingcart As an example the possibility of developing a porting

system for large-sized facilities like a shopping mall oran airport is studied in [4 5] To achieve this purposethey have attached a drive system to a shopping cart inorder to transmit power to the cart wheels To performthe desired person following capability they use a stereocamera installed in front of the cart One advantage of thissystem is that the person does not have to wear any attachedelement In contrast as they are using recognition by coloursegmentation the person has to wear clothes of a specificcolour Additionally they have to face all the problems relatedto computer vision high computational cost light variabilityproblems background colour interference problems and soforth

Amore complete prototype is proposed in [6] composedof two robots a companion and a carrier In order to localizethe position of the companion a set of cameras is used whichallows the detection of robots and humans present in thescenario The carrier robot is able to follow the companionone which is able to guide the customer to the desiredposition as well as follow him while waiting for the nextaction to be performed A system for tracking shopping cartsin indoor environments is developed in [7] This systemconsists of the installation of a computer and a video cameraon the shopping carts for them to be able to self-localizeand send their position to a centralized system We find thissystem interesting as it could also be a first step to having anautonomous cart with person following capabilities but as adisadvantage it requires a complex system to be installed onthe cart and a building with preexistent infrastructure

The study carried out in [8] concludes that elder peopleinteract in a better way with robots carrying the shoppingbasket when they have a humanoid aspect and provideconversational capabilities although this fact implies a highcost and complex system In [9 10] a shopping help systemis developed able to obtain the shopping list from a mobiledevice through a QR code carry the shopping basket andshow at each moment which articles are on it and communi-cate with the supermarket computer system to inform aboutthe location of articles It uses a laser range finder sonar andcontact sensors (bumpers) to navigate

A special case is presented in the system proposed in [11]where the user controls the robot remotely by specifying theproducts to be graspedThen the robot navigates through thesupermarket to perform the task autonomously A lot of workis provided in the design of the artificial hand used to grasp awide variety of food products No details are provided in thepaper about the localization and navigation techniques usedby the robot It seems a very useful system for handicappedpeople On the other hand some elderly people prefer beingpresent in the supermarket to touch the products and make amore detailed selection

23 Robot Person following Behaviour Person following per-formed by a robot (see Figure 2) has traditionally been animportant research topic in the machine vision area It isnatural to imagine that mobile robots of the future especiallythose operating in public places will be expected to have thisskill As an example in [12 13] two different efficient person-tracking algorithms for a vision-based mobile robot using
















ble1Com

paris


ofseveralreviewed

syste

ms

Sensors

Capabilities

Interaction

Camera

LRFlowast

RFID

Other

Follo

wing

Carryitems

Find

items

User

Environm

ent

Chen

andBirchfi

eld2007

[13]

Bino

cular

NO

NO

NO

NO

NO

NO

A

Chun

getal2012[15]

NO

NO

NO

NO

NO

NO

A

Cosgunetal2013[18]

NO

Kinect

na

NO

RemoteS

kype

chat

AFH


na

NO

BumpersQ

Rreaderson

arNO

NO

Disp

laykeyboard

FG

Germae

tal2009

[24]

Userd

etectio

nna

NO

NO

Touchpanel

F

GranataandBidaud

2012

[19]

Twocameras

NO

na

na

na

na

A

Grossetal200

9[1]

Omnidirectional

plus

frontal

NO

Sonarsensors

bumpers

NO

NO

Guide

user

totheitems

Touchpanel

speechhead

movem

ents

F

Kand

aetal2010

[2]

Bino

cular

NO

Person

identifi

catio

n

Speech

recogn

ition

floo

rsensors

NO

NO

Give

directions

Speecharm

and

head

movem

ents

H

Katic

andRo

dic2

010[3]

Inventorykeeping

Infrared

Hum

anassistance

AC

EFI

Kulyuk

inetal2005[23]

NO

Lo

calization

na

NO

Keypadspeech

D

Matsuhira

etal2010[6]

Stereo

plus

omnidirectional

NO

Touchandforce

sensorssonar

To

uchpanel

AFIJ

Nagum

oandOhya2

001[20]

NO

NO

UserL

ED

NO

NO

NO

A

Nish

imurae

tal2006

[4]

Stereo

Fu

ture

Encoder

NO

LEDs

AH

Shiehetal200

6[22]

Localization

na

na

LC

Dscreen

AD

Tomizaw

aetal2007

[11]

NO

na

NO

NO

Remotec

hat

FHK

Tsud

aetal2009

[14]

NO

NO

NO

Return

toorigin

NO

A

Yoshim

ietal2006

[21]

Bino

cular

NO

NO

Ultrason

ic

na

Touchpanel

speech

AF

Zend

eretal2007[16]

NO

NO

Odo

metry

na

NO

Speech

A

Zimmerman

2006

[7]

Mon

ochrom

atic

NO

NO

Barcod

efor

localizo

dom

Noautono

m

movem

ent

Not

implem

ented

Disp

layforitems

list

B

Com

paRo

b[35]

NO

Obstacle


Identifyitems

UserR

Fsonar

NO

Smartpho

neta

blet

PCF

lowast

Laserrange

finder

lowastlowast

Optionalfeature

nainform

ationno

tavailable

Aobstacle

(and

user)d

etectio

nB

barcod

efor

localization

Cbarcod

efor

itemsidentificatio

nDR

FIDforlocalization

ERF


nFdataconn

ectio

nwith

warehou

se

Gdatac

onnectionwith

user

Htele

operation

Imutualinteractio

nwith

otherrob

ots

Jenvironm

entcam

eras

forlocalization

Kob

jectmanipulation









P(x y)

d1 d2

dr







dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer


t1

t2

d



Rx radio module




119903

Measurementunits




1and 119862




1199032 0) and



(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)


119909 =1198892

1minus 1198892

2

2119889119903

119910 =


2119889119903

(2)



P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation
























ble1Com

paris


ofseveralreviewed

syste

ms

Sensors

Capabilities

Interaction

Camera

LRFlowast

RFID

Other

Follo

wing

Carryitems

Find

items

User

Environm

ent

Chen

andBirchfi

eld2007

[13]

Bino

cular

NO

NO

NO

NO

NO

NO

A

Chun

getal2012[15]

NO

NO

NO

NO

NO

NO

A

Cosgunetal2013[18]

NO

Kinect

na

NO

RemoteS

kype

chat

AFH


na

NO

BumpersQ

Rreaderson

arNO

NO

Disp

laykeyboard

FG

Germae

tal2009

[24]

Userd

etectio

nna

NO

NO

Touchpanel

F

GranataandBidaud

2012

[19]

Twocameras

NO

na

na

na

na

A

Grossetal200

9[1]

Omnidirectional

plus

frontal

NO

Sonarsensors

bumpers

NO

NO

Guide

user

totheitems

Touchpanel

speechhead

movem

ents

F

Kand

aetal2010

[2]

Bino

cular

NO

Person

identifi

catio

n

Speech

recogn

ition

floo

rsensors

NO

NO

Give

directions

Speecharm

and

head

movem

ents

H

Katic

andRo

dic2

010[3]

Inventorykeeping

Infrared

Hum

anassistance

AC

EFI

Kulyuk

inetal2005[23]

NO

Lo

calization

na

NO

Keypadspeech

D

Matsuhira

etal2010[6]

Stereo

plus

omnidirectional

NO

Touchandforce

sensorssonar

To

uchpanel

AFIJ

Nagum

oandOhya2

001[20]

NO

NO

UserL

ED

NO

NO

NO

A

Nish

imurae

tal2006

[4]

Stereo

Fu

ture

Encoder

NO

LEDs

AH

Shiehetal200

6[22]

Localization

na

na

LC

Dscreen

AD

Tomizaw

aetal2007

[11]

NO

na

NO

NO

Remotec

hat

FHK

Tsud

aetal2009

[14]

NO

NO

NO

Return

toorigin

NO

A

Yoshim

ietal2006

[21]

Bino

cular

NO

NO

Ultrason

ic

na

Touchpanel

speech

AF

Zend

eretal2007[16]

NO

NO

Odo

metry

na

NO

Speech

A

Zimmerman

2006

[7]

Mon

ochrom

atic

NO

NO

Barcod

efor

localizo

dom

Noautono

m

movem

ent

Not

implem

ented

Disp

layforitems

list

B

Com

paRo

b[35]

NO

Obstacle


Identifyitems

UserR

Fsonar

NO

Smartpho

neta

blet

PCF

lowast

Laserrange

finder

lowastlowast

Optionalfeature

nainform

ationno

tavailable

Aobstacle

(and

user)d

etectio

nB

barcod

efor

localization

Cbarcod

efor

itemsidentificatio

nDR

FIDforlocalization

ERF


nFdataconn

ectio

nwith

warehou

se

Gdatac

onnectionwith

user

Htele

operation

Imutualinteractio

nwith

otherrob

ots

Jenvironm

entcam

eras

forlocalization

Kob

jectmanipulation









P(x y)

d1 d2

dr







dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer


t1

t2

d



Rx radio module




119903

Measurementunits




1and 119862




1199032 0) and



(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)


119909 =1198892

1minus 1198892

2

2119889119903

119910 =


2119889119903

(2)



P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










ble1Com

paris


ofseveralreviewed

syste

ms

Sensors

Capabilities

Interaction

Camera

LRFlowast

RFID

Other

Follo

wing

Carryitems

Find

items

User

Environm

ent

Chen

andBirchfi

eld2007

[13]

Bino

cular

NO

NO

NO

NO

NO

NO

A

Chun

getal2012[15]

NO

NO

NO

NO

NO

NO

A

Cosgunetal2013[18]

NO

Kinect

na

NO

RemoteS

kype

chat

AFH


na

NO

BumpersQ

Rreaderson

arNO

NO

Disp

laykeyboard

FG

Germae

tal2009

[24]

Userd

etectio

nna

NO

NO

Touchpanel

F

GranataandBidaud

2012

[19]

Twocameras

NO

na

na

na

na

A

Grossetal200

9[1]

Omnidirectional

plus

frontal

NO

Sonarsensors

bumpers

NO

NO

Guide

user

totheitems

Touchpanel

speechhead

movem

ents

F

Kand

aetal2010

[2]

Bino

cular

NO

Person

identifi

catio

n

Speech

recogn

ition

floo

rsensors

NO

NO

Give

directions

Speecharm

and

head

movem

ents

H

Katic

andRo

dic2

010[3]

Inventorykeeping

Infrared

Hum

anassistance

AC

EFI

Kulyuk

inetal2005[23]

NO

Lo

calization

na

NO

Keypadspeech

D

Matsuhira

etal2010[6]

Stereo

plus

omnidirectional

NO

Touchandforce

sensorssonar

To

uchpanel

AFIJ

Nagum

oandOhya2

001[20]

NO

NO

UserL

ED

NO

NO

NO

A

Nish

imurae

tal2006

[4]

Stereo

Fu

ture

Encoder

NO

LEDs

AH

Shiehetal200

6[22]

Localization

na

na

LC

Dscreen

AD

Tomizaw

aetal2007

[11]

NO

na

NO

NO

Remotec

hat

FHK

Tsud

aetal2009

[14]

NO

NO

NO

Return

toorigin

NO

A

Yoshim

ietal2006

[21]

Bino

cular

NO

NO

Ultrason

ic

na

Touchpanel

speech

AF

Zend

eretal2007[16]

NO

NO

Odo

metry

na

NO

Speech

A

Zimmerman

2006

[7]

Mon

ochrom

atic

NO

NO

Barcod

efor

localizo

dom

Noautono

m

movem

ent

Not

implem

ented

Disp

layforitems

list

B

Com

paRo

b[35]

NO

Obstacle


Identifyitems

UserR

Fsonar

NO

Smartpho

neta

blet

PCF

lowast

Laserrange

finder

lowastlowast

Optionalfeature

nainform

ationno

tavailable

Aobstacle

(and

user)d

etectio

nB

barcod

efor

localization

Cbarcod

efor

itemsidentificatio

nDR

FIDforlocalization

ERF


nFdataconn

ectio

nwith

warehou

se

Gdatac

onnectionwith

user

Htele

operation

Imutualinteractio

nwith

otherrob

ots

Jenvironm

entcam

eras

forlocalization

Kob

jectmanipulation









P(x y)

d1 d2

dr







dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer


t1

t2

d



Rx radio module




119903

Measurementunits




1and 119862




1199032 0) and



(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)


119909 =1198892

1minus 1198892

2

2119889119903

119910 =


2119889119903

(2)



P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation

















P(x y)

d1 d2

dr







dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer


t1

t2

d



Rx radio module




119903

Measurementunits




1and 119862




1199032 0) and



(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)


119909 =1198892

1minus 1198892

2

2119889119903

119910 =


2119889119903

(2)



P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










dr d1

d2

P(xy)

CompaRob

CompaRob-2

TDoA

Ultrasonicring

Personfollowing

Tagreader

Tabletcomputer


t1

t2

d



Rx radio module




119903

Measurementunits




1and 119862




1199032 0) and



(119909 +119889119903

2)

2

+ 1199102

= 1198892

1

(119909 minus119889119903

2)

2

+ 1199102

= 1198892

2

(1)


119909 =1198892

1minus 1198892

2

2119889119903

119910 =


2119889119903

(2)



P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










P(x y)

C1

C2

d1 d2


2 0)(a)

P(x y)

C1

C2

d1 d2


2 0)

(b)



1and

1198892



1and 119889

2are taken by


1199032 0) and (119889



1and 119889




119903) between




800

700

600

500

400

300

200

100

014 ∘

17 ∘






















(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation


























(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










(a) (b)


(a) (b)


Follow Wait_1

Wait_2Search









Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










Follow

Wait_1

Wait_2

P(x y)

P(x998400 y

998400)






Time (s)0 10 20 30 40 50 60 70

r_1r_2r_3

r_4r_5

08

1

12

14

16

18

2

22

24

Dist

ance

(m)








r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











r_1r_2r_3

r_4r_5

0

1

2

3

4

5

6

7

8

9

10

Dist

ance

(m)

Detection threshold

50 100 150 200 250 300 3500

Time (s)





d = 02

d = 05

d = 10

10 20 30 40 50 60 700

Time (s)

0

05

1

15

2

25

3

35

Dist

ance

(m)






d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










d = 02

d = 05

d = 10

0

05

1

15

2

25

3

35

Dist

ance

(m)

10 20 30 40 50 60 700

Time (s)







44 cm 29 cm

65

cm






25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










25cm 40 cm

15 cm

22 cm

24 cm

22 cm

(a)

29

cm

44 cm

10 cm

65 cm

22 cm

22 cm

(b)






1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5
















Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation














Acknowledgments


References














































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation

































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









Documents

Research Article CompaRob: The Shopping Cart Assistance Robotdownloads.hindawi.com/journals/ijdsn/2016/4781280.pdf · 2016. 2. 7. · e study carried out in [ ] concludes that elder