Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
#DataSmartSummit
SUMMIT ON DATA-SMART GOVERNMENT #DataSmartSummit
APPLYING MOBILITY DATA IN CITIES
MICHAEL SCHNUERLEData Officer
City of Louisville@LouDataOfficer
November 2017
Louisville and Waze:Applying Mobility Data in Cities
Michael Schnuerle, Data Officer
Harvard Civic Analytics Network
Summit on Data-Smart Government
November 7, 2017
• Your survey results
• Lots of talking about Waze
• So many slides!
• Smartphone group activity
• Questions and deeper dive
Agenda
Survey Results ● Session participant survey● Civic Analytics Member survey
Harvard Session Survey
Harvard Session Survey
Civic Analytics Network Waze Survey10 Responses
• Allegheny County, PA
• Boston, MA
• Chicago, IL
• Kansas City, MO
• Los Angeles, CA
• New York City, NY
• San Diego, CA
• San Francisco, CA
• San Jose, CA
• South Bend, IN
● Not a Waze Partner● Waze Partner● Storing Data
No one is really using the data
Comments: data is hard to work with, hard to socialize around city, lots of use case ideas (common road backups, emergency management awareness, public visualizations, business development, incident management, historical analysis, real time travel times, street investment impact, city fleet routing, pothole repair, traffic coordination, flow and density, hot-spots, signal effectiveness, event congestion, parking)
Louisville and Innovation● Office of Civic Innovation● Mayor Fischer● Importance of Innovation
Office of Civic Innovation - LouisvilleMayor Greg Fischer Created office in 2010
Grace SimrallChief of Civic Innovation
Ed BlayneyInnovation Project Manager
Michael SchnuerleData Officer
Mary HamptonSenior Data Scientist
Matt Gotth-OlsenInnovation Project Manager
Chris SeidtCivic Technology Manager
Daro Mott Performance Improvement
Data Driven Culture in Louisville
Use data to drive decisions and see opportunities for improvement.
Mayor Greg Fischer at Code for America Summit
Importance of InnovationMost gov employees focused on the daily work. Mayor’s culture encourages breakthrough and innovation work.
Civic Innovation department is run like a start-up:Connect the dots, partner, prove value, find future funding.Help departments work smarter.
Focus Areas:
1. Smart City/Smart Home2. Tech Enabled Public Health3. Social Innovation4. Maker/Hacker Community
Values: Data-driven compassion/ digital inclusion
Waze CCP Overview● Importance of Transit - data, citizens, money, value● CCP Program● Quick Use Cases
• Lots of good data, connect to smart city
• High importance to citizens
• Able to allocate city budget
• Huge value and outcomes possible
• Equity and inclusion opportunities
Importance of Transit for Govs
Data Officer| Bellarmine PresentationOffice of PerformanceImprovement and InnovationOPI2
Waze.com/ccp
• Smartphone navigation app
• Partners with cities across the World with 2-way data sharing - 500+
• We send road closures/construction to open data
• They give us near real-time traffic jams, alerts, problem reports
Connected Citizens Program (CCP)
• Traffic Studies• Signal re-timings, before and after
• Faulty Equipment• Looking at traffic to discover device issues
• Early Analysis• Innovation team can analyze and suggest
Current Waze Project Examples
• Hot Spot Analysis• Evaluation of the most congested corridors
• Building Templates• Putting the right information in the right format
• Real–Time Alerts• Monitoring of a key corridors
Current Waze Project Examples
CCP Data Sharing - from Govs● Road Closure data● USDOT Workzone Standard for Autonomous Vehicles
• Published as open data for all to use
• Updated Daily
• Closures, construction, sidewalks, utility work
• Ingested by Waze for all users
Road Closure and Construction Data
data.louisvilleky.gov/dataset/lane-closures-construction
• Useful for Autonomous Vehicles• Collaborating with other cities and states • Detailed lane and turn details added• Todd A. Peterson, Office of Transportation
Operations, Federal Highway Administration is leading effort.
• Waze is looking at improving this data too.
USDOT Road Closure Workzone Data Standard
CCP Data Sharing - from Waze● Alerts● Jams● Irregularities● Data use agreement
• Near real-time data feed for:• Jams/Congestion (passively collected for users)
• Alerts (actively submitted by users)
• Potholes
• Roadkill
• Cars stopped on road
• Irregularities - speeds on slower roads
What does it give us?
• Missing Signs
• Weather
• Accidents
• Cannot:• share Waze data format spec • share raw Waze data• your CCP data feed URL
• Can:• use and share data internally• share derivative works• share with other CCP partners
Data Use Agreement
Waze Tools● Live Map● Traffic View● App● Future projects
Waze sends us traffic jams, alerts, and user reports. You can see all of that combined on Waze's Live Map: www.waze.com/en-GB/livemap
Everything from Live Map, plus your defined road corridors, unsual traffic. Waze's Traffic View Map: www.waze.com/en-GB/trafficview
Waze App - iOS and Android
Group Activity! ● Let’s Hack Waze!● Waze reports and the data feed
1) Open Waze App - Drive to a Gas Station
2) Report Harvard Roadkill!
3) Watch the Reports Roll In...FYI - Working on citizen defined areas with IFTTT integration
Waze History in Louisville ● Participation● Saving Data● Processing Internally● Hackathon● Processing Externally - Replicate to Other Cities
● Louisville Metro joined in Sept 2015 (5th city)
● Ed Blayney took ownership in Civic Innovation, data saving to Amazon AWS in June 2016
● Worked with IT to set up processing into internal MS SQL database, finished February 2017
● Ed working with Traffic to find more value and benefits. Socializing across city, understanding use cases, providing access/tools/training, defining data structures, merging other data (eg, hackathon)
● Michael Schnuerle building traffic data warehouse with Amazon City on Cloud grant in August 2017, Waze first data set.
● Collaborating with Waze, USDOT, cities for AWS replicable solution.
History of Waze partnership
Use Cases - Maps ● Event analysis● Corridor analysis● Citizen reports
Post Event Analysis
Thunder Over Louisville
Fireworks show to kick off two weeks of Kentucky Derby festivities
Animated Map
Change in Jams after re-timing
Pothole Reports - 311 vs Waze
Waze use is higher on faster roads vs 311 use on slower roads.
Static Map
Animated Map
Use Cases - Traffic Engineers ● Traffic Study Comparison● Hot-spot analysis● Inductive Loop● Other departments
● Traffic department uses Waze data over a time period to do a traffic
study on a corridor, instead of or in addition to a paid study.
● Traffic department monitors changes to a corridor based on tweaks, like
a lightweight traffic study.
● Traffic uses it for hot spot analysis and how to focus their resources.
● Traffic department can detect faulty non-connected equipment at
intersections (eg. induction loops)
● Traffic could use it for post event analysis to check effectiveness of road
closures and routing to optimize future events.
Current Traffic Use Cases
Define geographic corridors, look at jams and irregularities
Top 12 Worst Corridors
Hot Spot Analysis
● Police can monitor a reported road to check for speeders to
deploy officer strategically
● Police can point idle cameras to the location of a reported
collision and alert our Real Time Crime Center for quicker
situational awareness
● Public works can ingest pothole, downed trees, missing sign
reports, can compare to 311 calls
○ Waze use is higher on faster roads vs 311 use on slower roads.
Other Department In-Progress Use Cases
● Real-time alerts and routing for emergency vehicles
● Real-time collision user reports● Vision Zero analysis● Citizen facing notifications in custom
areas
Future Use Cases
911 ReportsGenesis PULSEMASSDOT
Real-time notifications of crashes and road closures
Ability to find crash scenes easier and faster
Best navigation to incidents using Waze SDK
Reducing response times
Waze Event Day
Navigation
New Stadium Game Days - Atlanta Braves
Waze Future - Analytics Tools for CCP
Internal Data Processing● Waze Data Structure ● Power BI● Limitations● FYI: GeoEvent Processor
RAW JSON DATA
FROM WAZE
• Contains descriptions, value ranges, data types for all data
• Available only to CCP partners
Waze Official Specification
Link to file
Internal Database Data Dictionary
• Data Access and visualization tool
• What our traffic department uses to access Waze database and run reports
Microsoft Power BI
Data processed into Louisville internal database
Limitations of Current MSSQL Database
1. Accessible only internally - can’t easily build APIs/services
2. Optimization of indexes = slow
3. No JDBC connections allowed (IT security)
4. Limited to MSSS Management Studio or Power BI access
5. Replication to other cities limited
• _
ESRI GeoEvent Processor: Alerts, actions, real-time
Internal Waze Hackathon● Planning and Setup● Outcomes● Institutionalizing Innovation
• Think technology/data jam session• Except with nerds
• Intense & Time-bound• Typically last for a day or weekend
• Aim for minimum-viable solutions• Not looking for perfect, just good enough
• Even just a good idea will do
What is a hackathon?
Waze Hackathon at LouieLab (city co-working space)
• 9am - Welcome - Ed Blayney, Michael Schnuerle, Mayor Fischer link
• 9:15 - Overview and Technical Details - Michael Schnuerle and Mike Reynolds
• 9:30 - GIS GeoEvent Processor and Waze/TARC - Andrew McKinney
• 9:40 - Current projects, new ideas, and pitches - Michael S, Ed B
• 10 - Organize into teams based on projects, start working with data - participants
• 12 - BYO Lunch or get Subway
• 2 - Finish projects, start group presentations, discuss successes and roadblocks, decide on future project work - Michael S, Ed B
• 3 - End of event, networking
Waze Hackathon Agenda
Waze Internal : Results• Traffic- Crash Analysis:
• Analysis of significant collisions on Waze
• Police- Cameras for Collisions:
• RTCC/ GeoEvent Processor Integration
• Police- Speeding Monitoring:
• Targeted speed enforcement/review from citizens
• Pollution- Heavy Truck Traffic:
• Air quality/heavy truck traffic
• Pollution- Getting Cars Off the Road:
• Pushing more people multi-modal uses
• Rain’s Effect on Driving:
• Driving habits/outcomes during rainfall events
BLOG POST
• Cast a wide net• Invite a variety of experts
• Diversity in thought leads to innovation
• Clear & Consistent Communication• Build a website with resources
• Let people know what to expect
• Get the data ready• Prepping the data allows for more action
• Learn what people want
Tips for hosting your own hackathon
• Strong Facilitation• Participants can get lost in the fray
• Help them stay on track
• Give them the insider knowledge they need
• Follow-up• Hackathons start ideas, they don’t finish them
• Good ideas should be followed up
• Celebrate success
Tips for hosting your own hackathon
Traffic Data Management ● Smart City Framework● Transit Data Warehouse● Amazon AWS● Cloud Benefits and Replication
Smart City Framework
Traffic Data Warehouse• Data Inventory
• Inventory of internal and external agencies transportation data
• Data Warehouse
• Technical build of the Data Warehouse in AWS
• Data Integrations
• Integrations of the various data-sets into the Data Warehouse
AWS City on a Cloud● Louisville won in the “Dream
Big” category this year.● $50K AWS credits for a
Traffic Data Warehouse● Waze is our first data set to
ingest
Waze Cloud Replication ● Collaboration on solution● AWS Architecture● Progress and Future
AWS Solution Collaboration1. Louisville is looking to create an automated AWS Waze
processing solution that can be replicated by any CCP Partner
2. You grab Cloud Formation code and deploy infrastructure stack to AWS
3. Enter your CCP data feed URL4. Analyze, query, extract live and historic data for your city5. We are working to be serverless and inexpensive, but
don’t know costs yet
Working Architecture - 1
Working Architecture - 2
Working Architecture - 3
Current AWS Cloud Formation1. Processes Waze CCP feed
every 2 minutes.2. Takes parameters:
EnvironmentName (dev, test, prod), S3ArtifactsBucket, WazeDataHttpUrl (CCP feed)
3. On Git: github.com/LouisvilleMetro/WazeCCPProcessor
USDOT Secure Data Commons ● ITS JPO Overview● Partnership● SDC Architecture● Informational Session
INVESTING IN A SHARED SOLUTION
SDC STARTING WITH WAZE DATA
Secure Data Commons Overview● Secure Data Commons (SDC) is a new initiative of Intelligent Transportation
System (ITS) Joint Program Office (JPO) from USDOT
● The vision of the ITS JPO Data Program is to increase the agility of the
transportation system to enable technology- and data-driven safety and
mobility gains.
● SDC will make it easy for USDOT and the broader transportation sector to
share and collaborate around sensitive data sets using modern, commercially
available tools that keep up with the pace of change in technology
● Inspired by how the health sector, specifically cancer researchers and
NIH/NCI, are taking on a similar challenge
Secure Data Commons Overview● SDC uses a new approach to provide secure, cloud-based, revocable access to
complex (high volume, velocity, and/or variety) and sensitive (containing private or
proprietary information) datasets in a secure enclave along with analysis tools and/ or
shared computing resources
● The SDC POC is intended as a test case for setting up a laboratory environment
concept for such datasets, and to generate technical, institutional, and lessons learned
that will inform the longer-term ITS JPO data governance strategy as well as other
USDOT efforts in this area
● The intended audience for the POC are researchers (from government, academia, and
industry), modal partners, technical support contractors, and private sector data
sources interested in accessing and using ITS-relevant transportation data that would
otherwise be unavailable
SDC Architecture1. Designed based on
Secure Cloud Computing Architecture (SCCA) from Defense Information Systems Agency (DISA)
2. Sophos provides VPN/firewall/IDS/IPS capabilities acts as a single point of entry and exit)
3. User access is secured through standalone Active Directory installed and configured within SDC Account
4. Connectivity between different VPC’s is through VPN gateways from VDSS to each of the VPC’s
5. Provision workstations for researchers using AWS App Stream
SDC Datalake Logical Architecture
Kibana Metadata Dashboard (development / draft)
Working AWS Architecture Diagram
USDOT Waze Coverage Area (development / draft)
USDOT vs Waze vs Your Datastore1. USDOT and Waze: similar to a vendor system
a. Good for some use cases and first pass analysisb. Targeted feature set with specific goalc. Rolls out updates based on client requests
2. Your custom cloud/internal datastorea. Extensible, exportable, flexible, customizableb. Detailed analysis, Additional data layer combinationsc. Integration with gov services and systemsd. Needs internal tech skills and time and cost
Conclusion ● Collaboration● Contact Information● Slide Deck LInk
Let’s Work Together!1. Leverage your city’s AWS
experts and collaborate on GitHub with us.
2. github.com/LouisvilleMetro/WazeCCPProcessor
3. Take the code and deploy it to AWS and give us feedback.
Michael Schnuerle
Data Officer
Twitter: @LouDataOfficer
Any Questions?
Contact Info:Michael Schnuerle
Data Officer
@LouDataOfficer
This Slide Deck:
LSVLL.io/LouisvilleWaze
Some content from Ed Blayney, Innovation PM at OPI -->