Business I ntelligence Development at Winnipeg Transit Bill Menzies - - PowerPoint PPT Presentation

Business I ntelligence Development at Winnipeg Transit

I TS Canada Webinar

February 28, 2013

Bill Menzies

Senior Transit Planner, Dillon Consulting Limited Manager of Service Development, Winnipeg Transit (formerly)

3

Service Plan Service Delivery Consolidation Analysis Information Decision Action

Data

Transit Service Management

4

APC

TSAS Database

Data Consolidation for Analysis

Geography Schedule Service Costs Fleet Dispatch Operations Utilization

GI S

Transit Cost Model

Network, Schedule, Cost Files (SAS) Bus Run Assignment Files (SAS) APC Files (DB2)

MMI S On-TRAC Contact NETData

6

Transit Network Definition

• Booking
• Route
• Route Direction
• Trip Pattern (Leg)
• Segment
• Defined by contiguous

timing points

• Common stop sequence
• Bus Stop
• Service Access/Egress
• Timing Point
• Transfer Point
• Relief Point

10

Utilization: APC System

• Automatic Passenger Counting System
• Boardings/Alightings by Stop
• Loads Between Stops
• Arrival/Departure/Dwell Times at Stops and Timing Points
• Actual Running Times Between Timing Points
• Service Delays
• 183 buses equipped (33% of fleet)
• 28 x 40’ High Floor (New Flyer D40)
• 150 x 40’ Low Floor (New Flyer D40LF)
• 5 x 30’ Low Floor (New Flyer D30LF)
• 3 service lines at 2 garages equipped

14

Data Consolidation

• Goal is to consolidate key data from different

systems to support analysis, planning, and management of transit service

• What’s needed:
• Data interfaces for major applications
• Software that can read a variety of database

formats

• End user programming skills
• Internal computer network for data access
• Intranet site for report publication

15

The TSAS Warehouse

Network & Schedule Database Bus/Run Assignments Database APC BusStops Database Service Cost Database Trip Samples Trip Sample Status Run Sample Status Trip Recovery Status Utilization Trip Productivity Run Times

SAS Programs

APC Diagnostic Reports Service Supply Reports Dispatch Reports Passenger Count Reports Service Evaluation Reports Schedule Adherence Reports Run Time Data Files for HASTUS-ATP

Performance Metrics

18

System-Based Metrics

• Dispatch
• % of scheduled service

actually operated

• % of designated service
• perated by designated

buses

• Low Floor Routes
• Downtown Spirit Routes
• Bike Rack Routes
• BRT Routes
• APC Data Recovery
• % of APC-operated trips for

which data successfully recovered

• Schedule Adherence
• % of departures within

window:

• 1 min early to 3 min late
• Measured only at stops with

boardings

• Weighted by boardings
• System Ridership
• Average Daily Boardings by:
• Booking Type
• Schedule Type
• Year

19

Route-Based Metrics

• Schedule Adherence
• % of departures within

window:

• 1 min early to 3 min late
• Weighted by boardings
• Running Times/Speed
• Mean, Standard Deviation
• Crowding
• At trip maximum load point,

% of passengers:

• Seated
• Standing (comfortable)
• Standing (uncomfortable)
• Trip Productivity
• Average Boardings
• Boardings/Bus Hour
• Maximum Load/Seated

Capacity

• % of Seat-Kms used
• Average pgr trip length
• Service Frequency
• Scheduled Headway vs.

Demand – Based Headway

• at each route’s maximum

load point

• By schedule type, time

period, and route direction

20

Stop-Based Metrics

• Passenger Activity
• Ons/Offs/Load by:
• Booking
• Schedule Type
• Time Period
• Route
• Schedule Adherence
• % of departures within

window:

• 1 min early to 3 min late
• Weighted by each

passenger trip’s boardings

21

40 seats

Service Evaluation - Supply

Bus Hours = 0.75

1000 m 1000 m 1000 m 1000 m 1000 m

45 Min 0 Min

Bus Kms = 5 Seat-Kms = (40 seats * 1,000 m)/ 1,000 m) * 5 = 200

22

40 seats

Service Evaluation - Demand

5 20 25 45 15 5 17 2 8 3 22 2 30 15

1000 m 1000 m 1000 m 1000 m 1000 m

45 Min 0 Min

Ons = 52 Offs = 52 Max Load = 45 Pgr-Kms = ( (5* 1000)+ (20* 1000)+ (25* 1000)+ (45* 1000)+ (15* 1000))/ 1000 = 110

23

Service Evaluation - Metrics

40 seats

5 20 25 45 15 5 17 2 8 3 22 2 30 15

1000 m 1000 m 1000 m 1000 m 1000 m

45 Min 0 Min

Boardings/ Bus Hour = 52/ 0.75 = 69 Boardings/ Bus Km = 52/ 5 = 10.4 Pgr-Kms/ Bus Kms = 110/ 5 = 22 (Average Load) Pgr-Kms/ Seat Kms = 110/ 200 = 55% (Load Factor) Pgr-Kms/ Boardings = 110/ 52 = 2.12 (Avge Pgr Trip Length)

24

Service Evaluation - Metrics

• Balanced Scorecard
• Performance Measures
• Boardings/Bus Hour
• Boardings/Bus Km
• Average Load
• Load Factor
• Complete Route
• Peak Direction
• Average Trip Length
• Cost Measures
• Variable Cost/Boarding
• Full Cost/Boarding
• Variable Cost/Bus Hour
• Full Cost/Bus Hour
• Route Peer Comparison
• Routes compared

within category:

• Downtown Main Line
• Express
• Crosstown
• Neighbourhood Feeder
• Downtown Spirit
• DART

25

Category Schedule Route Daily Service Operated Daily Costs and Demand Performance Measures Cost Measures

Service Evaluation - Example

I nformation Delivery

27

What, How, For Whom?

• On-Line Reports
• Generated daily at 05:00
• Designed for quick reference by all staff
• Detailed Paper Reports
• ~ 30 standard reports generated on demand for

planning/scheduling purposes

• Evaluation Reports
• Generated at end of each booking to track system ridership,

route productivity, demand-based headways, service costs

• Restricted circulation
• Data Extraction for Input to Other Applications
• Observed run time data for HASTUS-ATP analysis
• Spatial data for GIS analysis

28

On-Line Reports

29

System Schedule Adherence

30

Route Reports

31

Stop Reports

34

GI S – Load Profile

35

Next Steps

• Integrate data from new ITS deployments

into TSAS:

• Automatic Vehicle Location System (2010)
• Generates second-by-second log file of each bus’s
• peration each day
• Electronic Passenger Information Systems (2011)
• Trip planner, IVR, mobile, and SMS apps generate travel

pattern data

• Fare Collection System (2013)
• Generates spatially-referenced fare payment and transfer

data

• Develop management dashboard of key
• perational and performance measures

Thank You!