This document discusses a network-based approach to taxi sharing in New York City. It analyzes large GPS data sets of taxi movements to understand urban taxi systems and trips that could be combined. A new dispatch algorithm is proposed that can combine multiple trips into "Taxi Limousines" to make the system more efficient and reduce emissions. Simulations show that with only a 1 minute initial waiting time, most trips in NYC can be combined for taxi sharing. An online tool called HubCab is being developed to interactively explore the data.
2024.03.23 What do successful readers do - Sandy Millin for PARK.pptx
HubCab
1. Michael Szell mszell@mit.edu
Paolo Santi
Giovanni Resta
Stanislav Sobolevsky
Carlo Ratti
Steven Strogatz (Cornell)
Benedikt Groß
Joey Lee
Eric Baczuk
Carlo Ratti
Andi Weiß (47Nord)
Stefan Landsbeck (47Nord)
Research Visualization & Explorer
hubcab
Taxi-sharing in New York City: A network-based approach
2. Large GPS data sets on taxi movements
NYC
Singapore
13,500 cabs
26,000 cabs
Shanghai, San Francisco, Vienna, ...
3. Step 1: Analyze data
NY 170,000,000 trips / year
Pickups Dropoffs
7. Can we come up with a new system?
• More efficient
• Less emissions
• Affordable alternative
8. Step 2: A new dispatch algorithm
Combine 2 trips
9. Step 2: A new dispatch algorithm
Combine k trips “Taxi Limousine”
10. Manhattan street network
4000 intersections
9000 street segments
Extracted from
OpenStreetMap
Match GPS-coords of
pickup/dropoff points with
street intersections
11. Dynamic pickup and delivery problems
T1
T2
T3
T4
Like traveling salesman
with time constraints
Small systems solvable
with linear programming
Large systems not
Yang, Jaillet and Mahmassani, Transp Sci 38 (2004)
Berbeglia, Cordeau and Laporte, Eur J Op Res 202 (2010)
Marin, An Op Res 143 (2006)
13. Shareability networks
k = 2
T1
T2
T3
T4
T2T1
T3
T4
Solution: maximum matching
Generalizable to k>2
but unfeasible for k>3
Chandra and Halldorsson, J Alg 39 (2001)
14. Satisfaction criterion
Maximum time delay Δ
Δ = 30 sec Δ = 60 sec
more tolerance = denser network = more sharing opportunities
Krings et al, EPJ Data Sci 1 (2012)
15. Oracle vs. Online
Oracle: omniscient,
best possible
T1
T2
Online: realistic,
constrained by
time window δ
δ
Set δ = 1min
16. Step 2: A new dispatch algorithm
• Send destination request (via app)
• Wait δ min
• Receive sharing options
• Trip may be prolonged up to Δ min
How it works:
Consequences:
• Less traffic = less pollution etc
• Split costs for customers
17. Step 3: Simulation results: MOST trips can be combined!
Only δ = 1 min initial waiting time needed!
18. Online tool for interactive exploration
http://hubcab.org
(in development)
20. Michael Szell mszell@mit.edu
Benedikt Groß
Joey Lee
Eric Baczuk
Carlo Ratti
Andi Weiß (47Nord)
Stefan Landsbeck (47Nord)
Research Visualization & Explorer
hubcab
Taxi-sharing in New York City: A network-based approach
Paolo Santi
Giovanni Resta
Stanislav Sobolevsky
Carlo Ratti
Steven Strogatz (Cornell)