Reinventing Urban Transportation and Mobility. Pascal Van Hentenryck University of Michigan Ann Arbor, MI

Transcription

1 Reinventing Urban Transportation and Mobility Pascal Van Hentenryck University of Michigan Ann Arbor, MI 1

2 Outline Motivation Technology enablers Some case studies The MIDAS Ritmo projet Conclusion 2

3 The First/Last Mile Problem 3

4 The First/Last Mile Problem 4

5 The Importance of Mobility Car ownership in the US best predictor of upwards social mobility The relationship between transportation and social mobility is stronger than that between mobility and several other factors, like crime, elementaryschool test scores or the percentage of two-parent families in a community Nathaniel Hendren, Harvard University Transportation Emerges as Crucial to Escaping Poverty, New York Times, May

6 Congestion The cost of congestion in 2013, 124 billions predicted to be 184 billions in

7 The Challenge Can we transform mobility in a scalable way? 7

8 Outline Motivation Technology enablers Some case studies The MIDAS Ritmo projet Conclusion 8

9 Connectivity 9

10 Automated Vehicles 10

11 Progress in Analytics 11

12 Progress in Optimization If you only knew optimization from 10 years ago, you probably don t have the techniques needed to solve real-world sport scheduling problems Mike Trick, Professor at CMU, 2008 The following do make a big difference (and are much more recent ideas) Complicated variables Large neighborhood search Constraint programming (ideally combined with integer programming). 12

13 Outline Motivation Technology enablers Some case studies Public transportation in Canberra The MIDAS Ritmo projet Conclusion 13

14 Canberra 14

15 Planned City Garden city Walter Griffin Design principle self-contained communities greenbelt bush capital Many towns city centers infrastructure Started in

16 Urban Transportation The problem: off-peak bus service long routes 1-hour frequency buses running almost empty buses are expensive 16

17 Hub and Shuttle in Canberra 17

18 Urban Transportation 18

19 Public Transportation Descriptive Analytics bus boarding and alighting Discovering true O/D pairs from individual trips Predictive Analytics Predictive models for travel demand Prescriptive Analytics designing the network Benders decomposition online vehicle routing under uncertainty 19

20 Hub and Shuttle Transportation 20

21 Nature of the Trips 21

22 Case Study 22

23 Outline Motivation Technology enablers Some case studies Public transportation in Canberra Evacuation Planning The MIDAS Ritmo projet Conclusion 23

24 Rush Hours 24

25 Saturday Afternoon in AA 25

26 Prescriptive Evacuations 26

27 Prescriptive Evacuations 27

28 Prescriptive Evacuations 28

29 Prescriptive Evacuations 29

30 Scheduling Evacuations 30

31 Scheduling Evacuations 31

32 Disaster Management 32

33 The Inputs Text 33

34 20 v s 0 Scheduling Evacuations Time-expanded evacuation graph 9:00 10:00 11:00 12:00 13: Waiting Going from 2 to B takes 1 hour A At most 10 vehicles per hour B 2 3 A B 10 A B Node is flooded 10 5 at 11:00 10 A B 5 10 A B vt 34

35 Scheduling Evacuations Large-scale optimization model that needs to be solved in real time 10h horizon 5min steps? 185 nodes 458 edges nodes edges 35

36 Outline Motivation Technology enablers Some case studies Public transportation in Canberra Evacuation Planning The MIDAS Ritmo projet project vision Conclusion 36

37 MIDAS 37

38 Project Vision On-Demand Multimodal Transportation System multiple fleets of vehicles buses, shuttles, cars, light-rail, bicycles, pedestrian on-demand address the first/last mile problem human-centered mobility one click to order and trip tracking congestion management and quality of service routing and dispatching traffic lights and lane priorities pricing differentiated service infrastructure optimization road and bridge condition optimization 38

39 Outline Motivation Technology enablers Some case studies Public transportation in Canberra Evacuation Planning The MIDAS Ritmo projet project vision Ann Arbor as a living mobility lab Conclusion 39

40 UM Living Mobility Lab 40

41 UM Transit System Some figures 50,000 commuting trips a day 7.4 millions a year 75% capacity utilization increasing congestion issues 41

42 UM Ann Arbor Campus 42

43 Bus Routes and Capacity 43

44 The Research Team Ceren Budak, Assistant Professor, School of Information. Amy Cohn, Industrial and Operations Engineering. Rebecca Cunningham, M.D., Emergency Medicine, Tawanna Dillahunt, School of Information. Robert Hampshire, Transportation Research Institute. Jerome Lynch, Civil and Environmental Engineering Jonathan Levine, Taubman College of Architecture and Urban Planning. Louis Merlin, Taubman College of Architecture and Urban Planning. Jim Sayer, Transportation Research Institute. Pascal Van Hentenryck, Industrial and Operations Engineering. Michael Wellman, Computer Science & Engineering. 44

45 UM Parking and Transportation 45

46 Information and Technology Services 46

47 MTC 47

48 Outline Motivation Technology enablers Some case studies Public transportation in Canberra Evacuation Planning The MIDAS Ritmo projet project vision Ann Arbor as a living mobility lab some early steps Conclusion 48

49 UM Mobility Data 49

50 UM Mobility Data Application for experiment filed in eresearch Development in collaboration with Information and Technology Services (ITS) Bill Burns (Manager, Mobile/Portal/Web) and Jane Zhao Tom Amerman (Director Application Development) Advanced Research Computing Brock Palen, Jeffrey Sica Location data Real time 50

51 UM Buses 51

52 UM Buses (Commuter North) 52

53 UM Buses 53

54 UM Buses (NE Shuttle) 54

55 AATA Buses 55

56 Census Data 56

57 UM Mobility Data 57

58 Outline Motivation Technology enablers Some case studies The MIDAS Ritmo projet Conclusion 58

59 Conclusions Living Mobility Lab unique experimental laboratory 50,000 transit trips and >30,000 car trips a day data-rich environment freedom to experiment Next generation urban transportation systems multimodal transportation system on-demand service for first/last mile mobility economy of scale and congestion management pricing optimization of the underlying optimization 59