Upload
others
View
18
Download
0
Embed Size (px)
Citation preview
Supervision of Banana Transports
by the Intelligent Container
Coldchain Manangement. 4th International Workshop,
27./28. September 2010, Bonn
Reiner Jedermann, University Bremen, IMSAS / MCB
Axel Moehrke, Dole Europe Import BVBA, Belgium
Autonomous logisticsSFB 637MCBMCB 2
The banana challenge
Introduction (by Axel Moehrke) Bananas can produce up to 800 Watt of
heat per ton during ripening
Hard to control after a certain point
Improve processes along the transport and supply chain by permanent monitoring of quality changes
Untimely ripening is the greatest risk
Can be triggered in the field, in the packing plant or during transport
Or by insufficient cooling / temperature changes
Full and accurate temperature supervision and control is a precondition for improvements in the banana chain.
Autonomous logisticsSFB 637MCBMCB 3
Online monitoring by the intelligent container
Real-Time remote monitoring of local temperature
deviations
Idea first presented 2006
Transfer project 2008 + 2009
Online Access
Focus on results from field tests
Outlook to future research
Sensor
Nodes
External
Communication
Gateway
Pre-Processing
Autonomous logisticsSFB 637MCBMCB 4
Outline
Set up for field test
Observed temperature deviations
Over length of container
Between different containers
Inside one box
Required sensor density
Intelligent data processing
The future of the intelligent container
Autonomous logisticsSFB 637MCBMCB 5
Installation in Costa Rica
Gateway
Autonomous logisticsSFB 637MCBMCB 6
Installation in Costa Rica
4 pallets equipped with 4
sensors each S
S
S
S
S
Tier 1
Tier 8
Air ducts
Foam Block
Autonomous logisticsSFB 637MCBMCB 7
External Communication
Forwarding data from sensors to web-server
Using the vessel’s email system
Web-
Server
WLAN
Satellite
Wireless
SensorsGateway
Internet
Email Server
ShipContainer
Autonomous logisticsSFB 637MCBMCB 8
Webinterface
Mote-ID Last message: Temp. Humidity Voltage
1 2009-09-23 14:00:02 14.79 °C 93.0 % 2.8 V
2 2009-09-23 14:00:02 13.96 °C 94.0 % 2.81 V
3 2009-09-23 14:00:02 14.34 °C 95.0 % 2.84 V
4 2009-09-23 14:00:02 13.69 °C 81.0 % 2.8 V
5 2009-09-23 14:00:02 14.36 °C 102.0% 2.86 V
6 2009-09-23 14:00:02 15.2 °C 82.0 % 2.86 V
7 2009-09-23 14:00:02 15.57 °C 100.0% 2.75 V
8 2009-09-22 02:00:02 15.3 °C 97.0 % 2.82 V
Autonomous logisticsSFB 637MCBMCB 9
Signal attenuation by the fruits
Distance 0.5 meter ⅓ of all links
completely failed
⅓ of all links was not available part of the time
⅓ of all links worked well most of the time
Partly compensated by network
New hardware?0 2 4 6 8 10 12
0
0.5
1
1.5
2
2.5
Pallet
19Pallet
14Pallet
8Pallet
1
Door
End
Refr
igera
tion U
nit
Gate
way
1 2 3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Distance to cooling unit in [m]
Dis
tance to flo
or
in [m
]
Packet Rate: >0.75 0.33< <0.75 <0.33
Water-containing goods hinder the radio communication of wireless
sensors @ 2.4 GHz
Autonomous logisticsSFB 637MCBMCB 10
The necessity for core temperature measurement
The internal sensors of the aggregate help only little to
estimate the banana temperature
4 6 8 10 12 14 16 18 20
14
16
18
20
22
24
26
28
Days in August 2010
Tem
pera
ture
in [°C
]
Pallet core reefer side
Pallet core door side
Pallet 11 mid
Air supply (Aggregate sensor)
Air return (Aggregate sensor)
Autonomous logisticsSFB 637MCBMCB 11
Temperature difference over length of container
Bananas are cooled down by the container
Large differences in time required to achieve 17 °C
10 12 14 16 18 20 22 24
14
16
18
20
22
24
26
28
Time in [days] in September 2009
Pa
llet
Co
re T
em
pe
ratu
re in
[°C
]
Reefer Side Container 1
Door Side Container 1
Reefer Side Container 2
Door Side Container 2
Air Supply
Reefer side
2.4 days
Door side
6.35 days
Door-end
Reefer end
Two containers of
same type and
year of manufacture
Autonomous logisticsSFB 637MCBMCB 12
Comparison of 4 Experiments
Large variations in age of containers
Newer equipment cools faster, but local variations
cannot be avoided
The hot-spot can be at the door end or somewhere in
the middle of the container
0 2 4 6 8 10 12
2008
2009(A)
2009(B)
2010
Age 9 years
Age 12 years
Age 12 years
Age 2 years
Time to cool down to 17 °C in [days]
Reefer End
Door End
Maximum
Autonomous logisticsSFB 637MCBMCB 13
Location of hot-spots
Horizontal cut through the container
Average pallet core temperature in tier 5 (1.25 meter above floor)
Pallet 11 in the middle / left side is the hottest, but neighbor pallet
on the right side almost normal.
0 2 4 6 8 10 1215.4
15.6
15.8
16
16.2
16.4
16.6
Palle
t 1
Palle
t 1
Palle
t 3
Palle
t 11
Palle
t 11
Palle
t 10
Palle
t 19
16.20°C
15.80°C
16.42°C
16.59°C16.53°C
16.21°C
16.11°C
August2010
Distance to cooling unit in [m]
Avera
ge tem
pera
ture
in [°C
]
Hot-spot
Left side
Right
side
Door
Air supply
Autonomous logisticsSFB 637MCBMCB 14
Required sensor density
4 Sensors are not sufficient
Pallet at aggregate and door side / lowest and highest tier
But where to place
the extra sensors?
10 12 14 16 18 20 22 24
14
16
18
20
22
24
26
28
Container BH
Time in Days in September 2009
Tem
pera
ture
[°C
]
Full Temperature Range
Range covered by corner sensors
Corners Sensors: [ 5 8 17 19 ]
Extra Sensors: [ 7 13 ]
0 2 4 6 8 10 120
0.5
1
1.5
2
2.5
Pallet19
Pallet14
Pallet8
Pallet1
Do
or
En
d
Co
olin
g U
nit
Ba
se
16.0
17.8
18.5
17.9
16.7
16.8
16.7
16.9
15.8
16.3
16.5
---
15.9
16.2
16.1
---
Distance to cooling unit in [m]
Dis
tan
ce
to
flo
or
in [
m]
Average Box Temperature in [°C]
Autonomous logisticsSFB 637MCBMCB 15
4 6 8 10 12 14 16 18 20
14
16
18
20
22
24
26
28
Days in August 2010
Tem
pera
ture
in [°C
]
Accuracy of temperature measurements
How accurate can we measure temperature inside a
banana box?
Cooling air flows through the boxes variations
Even loggers in similar positions (distance ~ 5 cm)
≈ 0.1 °C at end of transport
Up to 1 °C during cool down
Temperature
difference in
one Box
Autonomous logisticsSFB 637MCBMCB 16
Accuracy of temperature measurements
Pulp temperature measurement
Hurts the banana
Comparison: Logger temperature before opening of pallets
Pulp temperature Logger in average 0.05 °C too high (σ =
0.085 °C)
Pulp temperature in centre of box 0.2 °C … 0.4 °C higher than
side of box
Autonomous logisticsSFB 637MCBMCB 17
Innovation Alliance
New project starts in September 2010
13 partner companies and 6 research institutes
Bremer Institut für
Produktion und Logistik
an der Universität Bremen
Sea-Containers
(Bananas)
Truck-Transports
(Meat)
Software RFID
and Electronics
Federal Ministry of
Education and
Research
Ethylene
Sensor
Autonomous logisticsSFB 637MCBMCB 18
Intelligent data processing
Not only forward data, but pre-
process
Different algorithms (OSGi
Software-Bundles) on demand,
similar to App-Store for mobile
phones
Elements of the decision
support tool
Spatial interpolation of
temperature by Kriging
Prediction of future
temperature curve
Shelf life models
High
sensor
tolerance
Autonomous logisticsSFB 637MCBMCB 19
10 15 20 25 30 35
14.5
15
15.5
16
16.5
17
15
.14
°C
15
.83
°C
16
.07
°C
15
.80
°C
16
.05
°C
16
.27
°C
16
.20
°C
15
.73
°C
16
.15
°C
16
.06
°C
16
.21
°C 16
.50
°C
16
.56
°C
15
.42
°C
15
.56
°C
16
.57
°C
16
.53
°C 16
.85
°C
16
.35
°C
16
.59
°C
15
.68
°C
15
.87
°C
16
.38
°C
16
.11
°C
15
.82
°C
15
.84
°C
16
.42
°C
Cro
wn
Rot
Ro
tten
Fin
ge
r
Ro
tten
Fin
ge
r
Button Number
Te
mpe
ratu
re in
[°C
]
Relation temperature and quality
No relation between box temperature and defects per box
Further influence factors have to be checked
O2, CO2, Ethylene
Age at harvest
Micro biological load
Mechanical damages
Autonomous logisticsSFB 637MCBMCB 20
0 0.2 0.4 0.6 0.8 1
2008
2009(A)
2009(B)
2010
Average defective fingers per box
Rotten Fingers
Crown Rot
Anthracnose
Relation temperature and quality
Relation to average quality per container
But no statistical evidence
Only on container level
No relation inside one container
0 2 4 6 8 10 12
2008
2009(A)
2009(B)
2010
Time to cool down to 17 °C in [days]
Reefer End
Door End
Maximum
No data available
Autonomous logisticsSFB 637MCBMCB 21
Summary
The pallet core temperature can show large variations
Differences in cool-down time
Temperature is tricky to measure
The hot-spot can be anywhere, will be missed if only 4 sensors
Even inside one banana box ΔT ≈ 0.5 °C
Relation of quality and temperature needs further
evaluation
Other factors likes gases and biological variance
The system will help to reduce losses by unwanted
ripening
Accurate temperature monitoring is the basis for the further steps
Compensate different quality levels by FEFO planning
Autonomous logisticsSFB 637MCBMCB 22
The End
Thanks for your attention
www.intelligentcontainer.com
Dr. Ing. Reiner Jedermann
University Bremen, FB1Institute for Microsensors, -actors and –systems (IMSAS)
Otto-Hahn-Allee, NW1
D-28359 Bremen, GERMANY
Phone +49 421 218 62603, Fax +49 421 218 98 62603
Email [email protected]
Jedermann, R: Autonome Sensorsysteme in der Transport-und Lebensmittellogistik, Dissertation Universität Bremen, Verlag Dr. Hut, 2009.
Autonomous logisticsSFB 637MCBMCB 23
Effects of interruptions of the power supply
Interruption of power supply during transshipment
(Harbor Costa Rica)
Only small rise
inside pallet
< 0.2 °C per hour
1d 1d6h 1d12h 1d18h 2d 2d6h 2d12h 2d18h 3d
14
15
16
17
18
19
20
21
22
23
24
25
Time in [Days/Hours] in May 2008
Tem
pera
ture
in [°C
]
Pallet Core
Pallet Surface Wall
Return Air Grid
Air Ducts Floor (Supply)