transit scheduling/routing data applications

50
Using Data to Design and Manage TriMet’s Route Network Portland State University Center for Transportation Studies Transportation Seminar Series November 2, 2012

Upload: trec-at-psu

Post on 02-Jul-2015

194 views

Category:

Documents


1 download

DESCRIPTION

Ken Zatarain / Bryan Rose, TriMet

TRANSCRIPT

Page 1: Transit Scheduling/Routing Data Applications

Using Data to

Design and Manage TriMet’s Route Network

Portland State University Center for Transportation

Studies Transportation Seminar Series

November 2, 2012

Page 2: Transit Scheduling/Routing Data Applications

Stop Records

Date

Line, Direction

Bus/Trip

Operator

Stop

Latitude/Longitude

Arrive/Leave Time

Dwell Time

Scheduled Leave Time

Passenger Ons/Offs

Page 3: Transit Scheduling/Routing Data Applications
Page 4: Transit Scheduling/Routing Data Applications

A HEAP of data….

Page 5: Transit Scheduling/Routing Data Applications

…needs to be structured for MANAGEMENT decisions

Page 6: Transit Scheduling/Routing Data Applications

When is a bus or train too crowded?

How do schedules account for variable traffic?

When and where should buses run more often?

How can street designs work better for transit?

Some Questions…

Page 7: Transit Scheduling/Routing Data Applications

Service Standards

Systematic, comprehensive, predictable method to design, evaluate and adjust service

Balance ridership, cost, community needs to improve usefulness of our service to the public

Page 8: Transit Scheduling/Routing Data Applications

Service Standards

Network Design Route structure and layout Policy headways Span of service Stop spacing

Page 9: Transit Scheduling/Routing Data Applications

Service Standards

Network Design Route structure and layout Policy headways & span of service Stop spacing

Service Priorities Passenger capacity Service reliability Frequency and span of service

Page 10: Transit Scheduling/Routing Data Applications
Page 11: Transit Scheduling/Routing Data Applications

Providing the Right Passenger Capacity

Page 12: Transit Scheduling/Routing Data Applications

Ridership Peaks

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

4 5 6 7 8 9 10 11 12p 1 2 3 4 5 6 7 8 9 10 11 12a 1 2

Spring 2012 Average Bus Boardings by Hour of the DayWeekday, Saturday and Sunday

Wkd Sat Sun

Page 13: Transit Scheduling/Routing Data Applications

More Peak Buses for Higher Loads

0

100

200

300

400

500

600

4 5 6 7 8 9 10 11 12p 1 2 3 4 5 6 7 8 9 10 11 12a 1 2

Fall 2012 Vehicles in Service by Hour of the DayWeekday, Saturday and Sunday

Wkd Sat Sun

Page 14: Transit Scheduling/Routing Data Applications

Efficient Addition of Service

Page 15: Transit Scheduling/Routing Data Applications

Schedule Efficiency

Leaves Powell

Garage at 7:11 am

Leaves Center

Garage at 5:24 am

8 min.First Stop at Holgate

and 134th at 7:19 am20 min.

First Stop at Holgate

and 134th at 5:44 am

64 min. 14 hrs 56 min.

Last Stop Downtown

Portland at 8:23 am

Last Stop Downtown

Portland at 8:40 pm

19 min.Returns to Powell

Garage at 8:42 am25 min.

Returns to Center

Garage at 9:05 pm

STATS STATS

InService Time 64 min. Report Time 15 min. InService Time 12h41m Report Time 15 min.

Pull Time 27 min. Pay Time 106 min Pull Time 0h45m Pay Time 15h56m

Layover/Recovery 0 min. Minimum Pay Time 120 min. Layover/Recovery 2h15m Minimum Pay Time 16h0m

Total Vehicle Hours 91 min In Service/Pay Time 53% Total Vehicle Hours 15h41m In Service/Pay Time 79%

Anatomy of a Tripper Anatomy of a Day Train

Page 16: Transit Scheduling/Routing Data Applications

Passenger Loading Standards

STANDARD 40 FOOT BUS Design Capacity = 64 Passengers Achievable Capacity = 51 Passengers (80% of Design Capacity) Seated Capacity = 39 Passengers STANDARD 2 CAR RAIL TRAIN Design Capacity = 322 Passengers Achievable Capacity = 266 Passengers (80% of Design Capacity) Seated Capacity = 132 Passengers

Benchmarks Prompting Consideration of Action

Current Signup Upcoming Signup

Bus 40% of trips >51 and/or 10%

> 64

>39 average (seated load)

Rail 40% of trips > 266 and/or

10% > 332

>213 average (80% of 266)

Peak (7-9a, 4-6p) Off-Peak

80% of trips > 266

and/or 20% > 332

>266 average

Current Signup Upcoming Signup

80% of trips >51

and/or 20% > 64

>51 average

Page 17: Transit Scheduling/Routing Data Applications

Loading Standards Applied Bus Load Analysis, Fall 2012 Signup (Through 10-08)

Top 25 Avg Loads - with loads of 51 or above highlighted

51 5 80% 20.0%

Rank Rte Dir Train Day Trip Start End Orig Dest Avg Load obs obs >51 obs >64 % >51 % >64 x

1 99 0 9906 W 1070 4:59 p 6:04 p Burnside Bridge Clackamas Community College 55 7 4 1 57% 14%

2 44 0 4405 W 1240 3:01 p 3:43 p North Terminal Westside PCC-\Sylvania 53 16 12 1 75% 6% x

3 35 0 3501 W 1090 7:11 a 7:51 a University of Portland SW 4th & Harrison 53 18 14 4 78% 22%

4 72 0 7249 W 1130 7:09 a 8:30 a Clac TC Max Station Anchor & Channel 52 10 7 0 70% 0% x

5 72 1 7239 W 1680 3:01 p 4:29 p Anchor & Channel Clac TC Max Station 52 16 8 3 50% 19% x

6 72 1 7258 W 1700 3:12 p 4:42 p Anchor & Channel Clac TC Max Station 52 14 9 0 64% 0% x

7 44 1 4406 W 1070 7:45 a 9:02 a PCC-\Sylvania Lombard & Pier Park 50 20 10 1 50% 5% x

8 12 0 1268 S 1360 6:08 p 7:15 p Parkrose TC 12 Bay Tigard\Transit Center 50 5 1 0 20% 0%

9 9 1 941 W 1230 8:28 a 9:40 a Gresham TC 9 Bay North Terminal Westside 50 14 4 1 29% 7%

10 61 0 6170 W 1030 4:30 p 5:11 p Veterans Rd & Bldg 16 Layover Beaverton TC 61 Bay 49 7 3 0 43% 0%

11 9 0 943 W 1390 3:00 p 3:50 p North Terminal Westside Powell Garage 49 20 8 0 40% 0%

12 71 1 7143 W 1090 7:30 a 9:02 a Clac TC Max Station Layover Foster at 94th 48 15 7 1 47% 7%

13 15 1 1535 W 1500 4:27 p 5:50 p Yeon & NW 44th Gateway Layover 48 22 8 1 36% 5%

14 15 1 1539 W 1560 5:09 p 6:30 p Yeon & NW 44th Gateway Layover 48 19 7 1 37% 5%

15 10 0 1004 W 1240 3:13 p 4:07 p Madison & 4th Layover Foster at 94th 48 13 4 0 31% 0%

16 9 1 901 W 1150 7:29 a 8:18 a Powell Garage North Terminal Westside 48 23 14 1 61% 4%

17 4 1 445 W 1560 2:30 p 4:39 p Gresham TC 4 Bay Richmond & Syracuse 48 13 5 0 38% 0%

18 99 0 9904 W 1040 4:01 p 5:05 p Burnside Bridge Clackamas Community College 47 19 6 0 32% 0%

19 75 1 7511 W 1400 3:20 p 5:05 p Lombard & Pier Park Milwaukie TC 21st/Jackson S 47 18 5 0 28% 0%

20 72 0 7234 W 1140 7:17 a 8:35 a Clac TC Max Station Anchor & Channel 47 12 6 0 50% 0%

21 72 1 7248 U 1310 1:27 p 2:46 p Anchor & Channel Clac TC Max Station 47 6 1 0 17% 0%

22 71 0 7138 W 1080 6:44 a 8:16 a Layover Foster at 94th Clac TC Max Station 47 14 6 0 43% 0%

23 35 0 3504 W 1320 4:26 p 6:00 p University of Portland Oregon City Transit Center 47 22 7 0 32% 0%

24 33 0 3308 W 1490 4:59 p 6:16 p on 5th NS Hoyt Clackamas Community College 47 24 9 0 38% 0%

BENCHMARKS

Page 18: Transit Scheduling/Routing Data Applications
Page 19: Transit Scheduling/Routing Data Applications

Writing Reliable Schedules

Page 20: Transit Scheduling/Routing Data Applications

On Time Performance

-3.00

-2.00

-1.00

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

4:48a 7:12a 9:36a 12:00p 2:24p 4:48p 7:12p 9:36p 12:00a

Min

ute

s Ea

rly

or

Late

On Time Performance Observations, Line 8 @ 5th and BroadwaySeptember 2 to November 3, 2012

SCHEDTIME

LATE

EARLY

Page 21: Transit Scheduling/Routing Data Applications

On Time Performance

80% 5 minutes late

1 minute early

Page 22: Transit Scheduling/Routing Data Applications

Run Time Varies through the Day

(new slide)

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

4:49

AM

5:08

AM

5:28

AM

5:48

AM

6:05

AM

6:23

AM

6:36

AM

6:49

AM

7:04

AM

7:18

AM

7:29

AM

7:39

AM

7:54

AM

8:06

AM

8:19

AM

8:36

AM

8:58

AM

9:18

AM

9:38

AM

9:58

AM

10:1

8 A

M

10:3

8 A

M

10:5

8 A

M

11:1

8 A

M

11:3

7 A

M

11:5

6 A

M

12:1

6 PM

12:3

6 PM

12:5

6 PM

1:15

PM

1:35

PM

1:55

PM

2:15

PM

2:30

PM

2:45

PM

3:00

PM

3:22

PM

3:43

PM

4:06

PM

4:29

PM

4:43

PM

4:57

PM

5:11

PM

5:30

PM

5:51

PM

6:12

PM

6:36

PM

6:56

PM

7:18

PM

7:38

PM

7:58

PM

8:18

PM

8:39

PM

9:00

PM

9:20

PM

9:50

PM

10:2

0 PM

10:5

0 PM

11:2

2 PM

11:5

2 PM

Scheduled Line 6-MLKRuntime from Jantzen Beach to 18th/Jefferson

Page 23: Transit Scheduling/Routing Data Applications

Run Time Varies Day to Day

Line 4 St. Johns to Gresham Run Times (PM Peak)

Page 24: Transit Scheduling/Routing Data Applications

What Run Times Should We

Use in the Schedule?

Page 25: Transit Scheduling/Routing Data Applications

What Run Times Should We

Use in the Schedule?

Page 26: Transit Scheduling/Routing Data Applications

Ranking Lines by Run Time Needs WEEKLY RUNTIME ANALYSIS (See Key at Bottom of Chart)

Line # 11TU 10TU 09TU 11TU 10TU 09TU 11TU 10TU 09TU

1 2 4 4 2.9% 5.8% 4.5% 4.2% 20.8% 19.2%

4 21 27 16 1.4% 1.8% 1.0% 17.4% 24.0% 7.9%

6 12 11 3 2.5% 2.3% 0.4% 10.6% 13.0% 1.4%

8 9 11 4 1.3% 1.7% 0.5% 3.9% 5.2% 1.1%

9 16 25 10 1.7% 2.7% 1.0% 16.6% 24.5% 5.7%

10 7 8 7 2.2% 2.5% 2.0% 5.5% 13.7% 7.4%

12 18 24 10 1.6% 2.1% 0.8% 16.3% 25.7% 9.6%

14 12 8 7 2.5% 1.7% 1.2% 4.9% 4.3% 1.9%

15 15 15 12 2.1% 2.0% 1.4% 12.3% 13.6% 4.4%

16 12 0 0 13.0% 0.2% 0.3% 63.6% 0.0% 0.0%

17 14 15 7 1.7% 1.9% 0.8% 14.2% 17.5% 6.6%

18 0 0 0 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

19 19 13 14 2.5% 1.7% 1.6% 20.2% 16.9% 14.7%

20 9 19 7 0.9% 1.9% 0.6% 6.7% 23.5% 7.3%

Weekly Hrs Needed to Bring All

Trips to 60th Percentile

Weekly Hrs Needed to Achieve

60th Percentile as % of Total Hrs

% of Trips That Need 3 Minutes of

Runtime to Achieve 60th Percentile

95th Percentile 16 95th Percentile 7.0% 95th Percentile 28.3%

80th Percentile 8 80th Percentile 3.9% 80th Percentile 13.7%

65th Percentile 4 65th Percentile 2.6% 65th Percentile 7.7%

Page 27: Transit Scheduling/Routing Data Applications

More traffic (auto, bike, pedestrian) Traffic calming

Bus Travel Time Impacts

Page 28: Transit Scheduling/Routing Data Applications

More traffic (auto, bike, pedestrian) Traffic calming + More ridership More diverse ridership Less frequency

Bus Travel Time Impacts

Page 29: Transit Scheduling/Routing Data Applications

More traffic (auto, bike, pedestrian) Traffic calming + More ridership More diverse ridership Less frequency =

Slower, more variable travel times

and higher costs

Bus Travel Time Impacts

Page 30: Transit Scheduling/Routing Data Applications

Run Time Comparison

Downtown Portland to Gresham

Page 31: Transit Scheduling/Routing Data Applications

Reliability Improvements

Page 32: Transit Scheduling/Routing Data Applications
Page 33: Transit Scheduling/Routing Data Applications

What Customers Want

Frequency

Reliability

Smooth ride

Shelter

Crowding

Seat available

Page 34: Transit Scheduling/Routing Data Applications

Customer Requests 2010-11

Frequency

Passenger crowding

Reliability

Route change

Unserved areas

Customers ask most for more frequent service

Page 35: Transit Scheduling/Routing Data Applications

Ridership Effectiveness Metric

Boarding rides per vehicle hour

Line

Day of week

Time period

Route segment

Vehicle hours = pull + revenue + layover hours

Page 36: Transit Scheduling/Routing Data Applications

Ridership Effectiveness by Line

ZEF spreadsheet example

Page 37: Transit Scheduling/Routing Data Applications

Ridership Effectiveness by Line

ZEF spreadsheet example

Page 38: Transit Scheduling/Routing Data Applications

Higher Load ≠ Higher Effectiveness

EffectivenessSimpler

Page 39: Transit Scheduling/Routing Data Applications

Estimating Ridership RIDERSHIP ESTIMATES FOR VARIOUS OPTIONS TO SERVE BARROWS LOOP

Feeder Coefficents MODEL CALIBRATION

Pop 0.010 - To keep the model balanced

Emp 0.004 - An adjustment of any coeficient in one direction requires

Ret 0.008 the adjustment of another in the opposite direction

- If the Difference (cell D7) is > 0 then the models

Difference 31 estimates are high, < 0 then estimates are low.

R-Squared 0.967746062 - The R Squared figure should be kept close to 1

# Pop Emp Retail Trips Estimate

Similar Routes w/o ret. emp. w/o trips Elasticity Estimate Actual

25 Glisan-Rockwood 21576 2208 6507 26 273 1.00 197 147

27 Market 20,020 3,920 8,355 24 280 1.00 186 139

29 Lake-Webster 13590 9265 12311 39 269 1.00 292 370

29 Lake 13,590 9,265 12,311 39 269 1.00 292 346

36 South Shore 28084 11323 13865 28 433 1.00 336 350

37 North Shore 8,891 8,667 10,512 23 206 1.00 132 118

46 North Hillsboro 10,568 8,531 2,876 40 161 1.00 179 168

47 Baseline / Evergreen 21,994 9,014 10,977 65 340 1.00 614 587

48 Cornell 17,600 11,001 9,842 63 296 1.00 518 502

59 Walker / Park Way 11,889 4,064 5,095 60 174 1.00 290 293

60 Leahy Road 3,964 726 8,591 31 111 1.00 95 42

154 Willamette 9506 2847 2912 28 128 1.00 100 140

# Pop Emp Retail Trips Estimate

Barrows Loop Estimates w/o ret. emp. w/o trips Elasticity Estimate Actual

Barrows Loop Estimates

62-Scholls-Horizon-Barrows 8400 130 900 70 90 1.00 176

62-Scholls-Horizon-Murray 5600 120 700 70 61 1.00 119

62-Scholls-Horizon-135th 9200 130 950 70 99 1.00 192

62-Barrows Loop 11900 220 1000 70 126 1.00 245

62-Barrows-Murray Loop 9300 167 800 70 99 1.00 192

56-Extension 7500 100 750 70 80 1.00 156

These Rows Need to Be Filled In Copy these rows down

These Rows Need to Be Filled In Copy these rows down

Page 40: Transit Scheduling/Routing Data Applications

Span of Service

Page 41: Transit Scheduling/Routing Data Applications

Improving Span of Service

Page 42: Transit Scheduling/Routing Data Applications

Improving Span of Service

Earlier Trip

Later Trip

Later Trip

Page 43: Transit Scheduling/Routing Data Applications

Ridership by Segment

Page 44: Transit Scheduling/Routing Data Applications
Page 45: Transit Scheduling/Routing Data Applications

Street Design for Transit

Analysis Slow or inconsistent travel times Delay hot spots Passenger boardings

Page 46: Transit Scheduling/Routing Data Applications

Street Design for Transit

Analysis Slow or inconsistent travel times Delay hotspots Passenger boardings

Service Design Stop spacing Shelter location and size Passenger amenities Transit priority treatments

Page 47: Transit Scheduling/Routing Data Applications

Street Design for Transit

Enhancements Stop consolidation Queue bypass lanes Leading green Signal priority Bus bulbs

Page 48: Transit Scheduling/Routing Data Applications

How Big Should Each Mall Shelter Be?

0

10

20

30

40

50

60

0

10

20

30

40

50

60 4:3

0 p

4:3

1 p

4:3

2 p

4:3

3 p

4:3

4 p

4:3

5 p

4:3

6 p

4:3

7 p

4:3

8 p

4:3

9 p

4:4

1 p

4:4

2 p

4:4

3 p

4:4

4 p

4:4

5 p

4:4

6 p

4:4

7 p

4:4

8 p

4:4

9 p

4:5

0 p

4:5

1 p

4:5

2 p

4:5

3 p

4:5

4 p

4:5

5 p

4:5

6 p

4:5

7 p

4:5

8 p

4:5

9 p

5:0

0 p

5:0

1 p

5:0

2 p

5:0

3 p

5:0

4 p

5:0

5 p

5:0

6 p

5:0

7 p

5:0

8 p

5:0

9 p

5:1

0 p

5:1

1 p

5:1

2 p

5:1

3 p

5:1

4 p

5:1

5 p

5:1

6 p

5:1

7 p

5:1

8 p

5:1

9 p

5:2

0 p

5:2

1 p

5:2

2 p

5:2

3 p

5:2

4 p

5:2

5 p

5:2

6 p

5:2

7 p

5:2

8 p

5:2

9 p

Sum 4 9 14 17

5th and Yamhill

Fall '04 Passenger Accumulation

Page 49: Transit Scheduling/Routing Data Applications

0

10

20

30

40

50

60

0

10

20

30

40

50

60

4:3

0 p

4:3

1 p

4:3

2 p

4:3

3 p

4:3

4 p

4:3

5 p

4:3

6 p

4:3

7 p

4:3

8 p

4:3

9 p

4:4

1 p

4:4

2 p

4:4

3 p

4:4

4 p

4:4

5 p

4:4

6 p

4:4

7 p

4:4

8 p

4:4

9 p

4:5

0 p

4:5

1 p

4:5

2 p

4:5

3 p

4:5

4 p

4:5

5 p

4:5

6 p

4:5

7 p

4:5

8 p

4:5

9 p

5:0

0 p

5:0

1 p

5:0

2 p

5:0

3 p

5:0

4 p

5:0

5 p

5:0

6 p

5:0

7 p

5:0

8 p

5:0

9 p

5:1

0 p

5:1

1 p

5:1

2 p

5:1

3 p

5:1

4 p

5:1

5 p

5:1

6 p

5:1

7 p

5:1

8 p

5:1

9 p

5:2

0 p

5:2

1 p

5:2

2 p

5:2

3 p

5:2

4 p

5:2

5 p

5:2

6 p

5:2

7 p

5:2

8 p

5:2

9 p

Sum 4 9 14 17

5th and Yamhill

Fall '04 Passenger Accumulation

How Big Should Each Mall Shelter Be?

Page 50: Transit Scheduling/Routing Data Applications

Bryan Rose – [email protected]

Ken Zatarain – [email protected]

Some More Questions…?