This was a "just for fun" project as part of the US DOT + Uber Hackathon. Feb, 2016.
The goal was to use Open Data from the government, as well as Uber's API, to help solve potential policy issues.
http://celebratingcities.github.io/
Note that we had 24 hrs to form a team, create a product, and present.
1. First/Last Mile
Problem Solving
The Trip Affordability Tool
An application of real-time data collection and economic analysis
Presentation for the Department of Transportation & Uber
Celebrating Cities - Hack the Last Mile
2. Objectives
• For someone with no computer skills:
✦ Learn Something
✦ Work to Make Something Useful
• Focus on Civic Utility
5. But Context is More Important
Source: MD iMap, Open Street Maps
6. Dedicated Transit Entry Points Provide
Access to Transportation Networks
Which Can Provide Access to Jobs
Source: Location Affordability Index, Transpo Cost as % of Income, Low Income Individuals
7. How do People Get to Entry
Points?
• Walk
• Bus
• Bike
• Rail
• Auto
8. How Do They Choose Which
Mode?
• And are they accurately taking into accounts all the costs of their trip?
• US DOT Guidance: The Value of Travel Time Savings: Departmental
Guidance for Conducting Economic Evaluations Revision 2 (2015 Update)
• “it should not be inferred that travelers prorate their household
incomes by the hour to make decisions.”
10. Take the LAI to a Micro-Level
Trip Decision Making
11. Product Development:
Start Small
• Commuter Trips are Often Multi-Segment
• Focus on 1st Piece
• Simple Calculations to Allow Person to See and
Understand Cost
12. Overview
• User enters address
• Fields return nearest Light Rail Station with additional information
• Time to Walk
• Cost of Walk Based on Value of Time
• Time for Uber Trip
• Cost of Uber Trip
• Allows user to see true cost
14. Technical Approach
• Python Jupyter Notebook (https://gist.github.com/marks/79517bbd95fb159174e7)
• Calculate Nearest Light Rail Station based on distance from User Starting Address
• Use Google Maps API to return Walk time information (and Bus, Rail, Other)
• Calculate Cost of Walk Based on Value of Time
• Use Uber API
• /V1/REQUESTS/ESTIMATE
• pickup_estimate
• price
• trip.duration_estimate
16. Limitations
• Tool is for One Segment of Trip
• Only Monetizes Time
• Certain modes have use-fees (fares,
depreciation, subscription, etc.)
• No weighting of values - people have different
preferences, which influence decisions
17. Future Applications
• User Can Optimize Mode Accessibility by Cost
• User Can take Measures to Reduce Mode Trip Cost
• Use Dynamic Location Placement for
Pickup/Dropoffs for Pooled Services