VTA HACKATHON Gather ideas for how to visualize and leverage real-time data Swiftly received multiple awards • • VTA established partnership with Swiftly VTA feeds real time data to Swiftly – – VTA used Swiftly data in the VTA Trip Planner VTA began utilizing Swiftly reports and – analytics tools
THE PROBLEM Hard to make use of messy and large volumes of data
FUTURE DATA FLOW Centralized data collection, analysis, visualization, and dissemination hub
DATA GOALS Create high quality data for internal and external use 1. 1. Ex Exte tern rnal use 1. Mobile apps 2. Web-based trip planners 3. Electronic stop signs 4. SMS & Voice 2. 2. In Inte tern rnal us use: Planning – – Scheduling – Operations
WHY THIS IS IMPORTANT Goal: A conceptual model for measuring transit induced stress at stops. Em Empa path thy Measure res: Visib Vis ible • – Amenities No Non-Vis Visib ible • – Schedule Adherence Trip Frequency – Da Data Sources: VTA Transit Passenger Environment Plan (TPEP) • • Swiftly on-time performance
External Use Real-time Passenger Information
PASSENGER EXPECTATIONS More Riders Happy Riders Time Savings 1.7% increase in Customers report 92% of customers NYC weekday report greater an average of 2 ridership minutes saved satisfaction
YOU MAY NOT KNOW 9% of riders say they reduced their transit use after receiving errors in real-time predictions. Source: “Benefits of Real-Time Transit Information and Impacts of Data Accuracy on Rider Experience” by Aaron Gooze, Kari Edison Watkins, and Alan Borning
OUR GOAL: DATA ACCURACY LEGACY CAD/AVL WIFI 2 minute polling rate 5 second polling rate Combine feeds in real-time • • Sophisticated prediction algorithms Up to 30% increase in RTPI • accuracy Stop Signs Etc… Mobile Apps Trip Planners
VTA TRIP PLANNER VTA Real Time data processed by Swiftly used in Trip Planner Real Time information of Vehicles Real Time information at Stops
Internal Use Analytics for Planning, Scheduling, and Operations
PERFORMANCE MONITORING • Very difficult to get clean data • Only can use CAD/AVL data which has data accuracy challenges Hard to analyze large datasets • PREVIOUS WORK FLOW NEW WORK FLOW Download Data Download Data Plot Data Analyze Data Segments Join Data Export Data Results Re-Join Import Data Re-Export Analyze Data Results
OUR GOAL: DATA ACCURACY & RAPID ANALYSIS LEGACY CAD/AVL WIFI 2 minute polling rate 5 second polling rate • Combine feeds in real-time Rapid big data analysis • On-Time Vehicle Dwell Times Etc… Performance Speeds
THE PLATFORM It’s all about computing and visualizing big data Clean and high fidelity data • • Analyze millions of data points in seconds Training staff • Transit Planners – Service Planners – – Reporting and Analysis Providing feedback •
ON-TIME PERFORMANCE Using Swiftly to discover performance issues
DRILL INTO THE DETAILS TIME OF DAY STOP LEVEL TRIP & STOP SEVERITY (HISTOGRAM)
CAUSES OF TRANSIT STRESS How can we use data and analysis to prioritize transit environment improvements • You Miss your Ride • Transit leaves Early • Transit arrives Late • Dark un-lit stop • Dirty stop • No ability to know when the next vehicle will arrive Tr Tran ansi sit pl planners rs and ope pera rato tors rs must t de demonstrate Empathy.
TRANSIT PLANNING: SPEEDS & DELAYS
DWELL TIME Stop dwell Stop dwell SANTA CLARA & 1ST 69.5 SANTA CLARA & 13TH 8.9 1ST & PASEO DE SAN ANTONIO 58.0 SAN CARLOS & DELMAS 8.7 VALLEY FAIR TRANSIT CENTER 55.2 STEVENS CREEK & LOPINA 8.7 STEVENS CREEK & KIELY 39.3 STEVENS CREEK & PORTAL 8.6 37.6 8.4 STEVENS CREEK & LOMA LINDA ALUM ROCK & MCCREERY SAN CARLOS & BIRD 36.5 STEVENS CREEK & ALBANY 8.1 ALUM ROCK & JACKSON 34.4 BELLEROSE & CLARMAR 7.7 29.3 7.6 SANTA CLARA & 3RD STEVENS CREEK & MAPLEWOOD STEVENS CREEK & MILLER 29.2 ALUM ROCK & 34TH 7.4 ALUM ROCK & KING 27.4 ALUM ROCK & MUIRFIELD 7.4 26.0 7.4 SAN CARLOS & BASCOM STEVENS CREEK & RICHFIELD SANTA CLARA & 17TH 25.8 SAN CARLOS & JOSEFA 7.0 25.6 6.7 SAN CARLOS & GRAND STEVENS CREEK & CASA VIEW SANTA CLARA & 7TH 22.2 ALUM ROCK & CHECKERS 5.9 STEVENS CREEK & STERN 21.7 STEVENS CREEK & HENRY 5.7
DWELL TIME: DEEPER DIVE
TRANSIT SIGNAL PRIORITY & DWELL TIME
STOP LEVEL ANALYSIS
TRANSIT NETWORK RE-DESIGN • Transit stress model identified areas that align with routes that have been slated for removal in the Next Network • Two independent studies providing similar results. Draft Plan Final Plan
SWIFTLY TECHNOLOGY ALLIANCE A forum to share learnings 8/30/17 • • CTA, Miami-Dade, Honolulu, MBTA, VTA, RATP Reviewed VTA analysis • Reviewed New Dashboard • • Continued development of Swiftly for improved Transit Big Data Analysis
PREGUNTAS? QUESTIONS? Thank you to: • – Jason Kim, Senior Transportation Planner, VTA Vivek Bansal, GIS Programmer, VTA – Mike Smith, CIO & Cofounder, Swiftly – – Will Dayton, CTO & Cofounder, Swiftly
Recommend
More recommend
