TAMING BIG DATA FOR SMARTER FUTURE

2015 APA National Planning Conference Function Code: S585 TAMING BIG DATA FOR SMARTER FUTURE Ed Hutchinson

OVERVIEW 1. Enterprise Information Management (EIM) 2. Data Clearing House 3. Florida Transportation Data Symposium

ENTERPRISE INFORMATION MANAGEMENT Enterprise view of maintaining and managing all areas related to management of the Department s data. Information Technology Strategic Initiatives Enterprise Information Technology Governance Enterprise Information Management Uniform Information Technology Standards

ENTERPRISE INFORMATION TECHNOLOGY GOVERNANCE The purpose of the Data Governance Document is to document FDOT s approach to governing how data will be stored, shared and protected within the organization.

DATA CLEARINGHOUSE Data Sources Data Users 26 MPOs 14 Seaports Private Sector FDOT Modal Offices 7 FDOT Districts

DATA CLEARINGHOUSE Data Sources Data Users 27 MPOs 14 Seaports Private Sector 7 FDOT Districts

DATA CLEARINGHOUSE Data Sources The addition of added Value Bottleneck Study Data Users 27 MPOs 14 Seaports Private Sector 7 FDOT Districts

DATA CLEARINGHOUSE

FLORIDA TRANSPORTATION DATA SYMPOSIUM!!!!! Save the Date August 18-20th 2015 Omni Orlando Resort at Champions Gate Twitter: @FLDATASYMPOSIUM Contact Email: FLTDS@dot.state.fl.us Promotion Website: http://fltds.com

CONTACT Ed Hutchinson FDOT Ed.Hutchinson@dot.state.fl.us 1-850-414-4910

2015 APA National Planning Conference Function Code: S585 TAMING BIG DATA FOR SMARTER FUTURE Ed Hutchinson

OVERVIEW!Enterprise Information Management (EIM)!Data Clearing House!Florida Transportation Data Symposium

ENTERPRISE INFORMATION MANAGEMENT!Enterprise view of maintaining and managing all areas related to management of the Department s data.!information Technology Strategic Initiatives "Enterprise Information Technology Governance "Enterprise Information Management "Uniform Information Technology Standards

DATA CLEARINGHOUSE Data Sources Data Users 26 MPOs 14 Seaports Private Sector FDOT Modal Offices 7 FDOT Districts

DATA CLEARINGHOUSE Data Sources Data Users 27 MPOs 14 Seaports Private Sector 7 FDOT Districts

DATA CLEARINGHOUSE Data Sources The addition of added Value Data Users 27 MPOs Bottleneck Study 14 Seaports Private Sector 7 FDOT Districts

FLORIDA TRANSPORTATION DATA SYMPOSIUM! Save the Date August 18-20th 2015! Omni Orlando Resort at Champions Gate Twitter: @FLDATASYMPOSIUM Contact Email: FLTDS@dot.state.fl.us Promotion Website: http://fltds.com!!!

CONTACT Ed Hutchinson FDOT Ed.Hutchinson@dot.state.fl.us 1-850-414-4910

USING DATA TO IMPROVE SAFETY Krishnan Viswanathan, CDM Smith

AGENDA Purpose Technology Study Design User Interface Data Processing Results Summary and Next Steps

PURPOSE Study Level 0 Automation Evaluate impacts of Advanced Driver Assistance System (ADAS) Measure changes in driver behavior in face of driver feedback technology Measure performance improvements as a result of feedback

TECHNOLOGY - GEOTAB Provides vehicle diagnostics information Real time data on vehicle exceptions and performance Data transmitted via cellular signals Dashboard to visualize data

TECHNOLOGY - MOBILEYE Daylight Pedestrian Collision Warning, including Bicycle Detection Forward Collision Warning, both in Highway and Urban areas, including Motorcycle Detection Lane Departure Warning Headway Monitoring and Warning

STUDY DESIGN Instrument 100 vehicles in Florida DOT District 7 Vehicles from Florida DOT and local transit agencies 100 vehicles have GeoTab devices 50 vehicles have Mobileye devices Study duration is 1 year Installs completed in August Data collection underway

USER INTERFACE

DATA PROCESSING Daily stream of data! Mobileye exceptions! Geotab exceptions Data arrives with lat/long information, time of day, and vehicle id for each type of Mobileye or Geotab exception Data converted into R data.table for processing and analysis ArcGIS used to spatially locate data

RESULTS ODOMETER READINGS Sep 2014 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 Min 47 68 13 104 1-1 Max 18,001 21,382 22,440 19,634 23,968 13,820 12,838 Median 3,896 4,049 3,189 3,216 3,388 3,203 2,765 Mean 4,461 4,812 4,301 4,301 4,414 4,274 3,863

RESULTS MONTHLY MILEAGE BY VEHICLE 30000 25000 Sep 2014 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 20000 15000 10000 5000 0

RESULTS DISTANCE, SPEED, TRIP DURATION

RESULTS MOBILEYE EXCEPTIONS Type of Warning Mobileye Events (September 2014) Mobileye Events (October 2014) Mobileye Events (November 2014) Mobileye Events (December 2014) Mobileye Events (January 2015) Mobileye Events Mobileye Events (February 2015) (March 2015) # % # % # % # % # % # % # % ME Forward Collision Warning 1,107 0.9% 1,034 1.0% 189 1.0% 179 1.0% 218 0.9% 180 0.8% 162 0.9% ME Headway Warning (Tailgating) 69,967 59.5% 65,401 65.4% 8,310 45.3% 7,547 42.2% 11,860 49.2% 10,077 46.3% 8,283 45.0% ME Left Lane Departure 14,124 12.0% 10,653 10.6% 1,743 9.5% 2,140 12.0% 2,233 9.3% 2,290 10.5% 1,815 9.9% ME Pedestrian Collision Warning 645 0.5% 433 0.4% 164 0.9% 144 0.8% 178 0.7% 127 0.6% 75 0.4% ME Right Lane Departure 29,617 25.2% 21,816 21.8% 7,504 40.9% 7,517 42.1% 9,069 37.6% 8,778 40.3% 7,728 42/0% ME Urban Forward Collision Warning 2,204 1.9% 724 0.7% 432 2.4% 345 1.9% 544 2.3% 338 1.6% 361 2.0% Total 117,664 100% 100,061 100% 18,342 100% 17,872 100% 24,102 100% 21,790 100% 18,424 100%

RESULTS MOBILEYE EVENTS BY VEHICLE TYPE 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% Car Large Bus Pickup Small Bus Truck Van ME Forward Collision Warning ME Headway Warning (Tailgating) ME Left Lane Departure ME Right Lane Departure ME Pedestrian Collision Warning ME Urban Forward Collision Warning

RESULTS GEOTAB EXCEPTIONS Sep-14 Oct-14 Nov-14 Dec-14 Jan-15 Feb-15 Mar-15 Speeding 19,380 23,176 17,707 16,966 16,846 14,448 16,517 Hard Acceleration 1,423 1,195 986 1,199 1,042 1,166 1,120 Harsh Cornering 570 877 622 757 717 688 906 Harsh Braking 106 264 125 242 268 110 194 Total 21,479 25,512 19,440 19,164 18,873 16,412 18,737

RESULTS COMPARISONS OF VEHICLES WITH AND WITHOUT MOBILEYE Anchor Towing FDOT D7 Car 120 100 80 150 100 60 40 50 20 0 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 0 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 Anchor Towing, Administrative, Pickup - Standard Anchor Towing, Administrative, Pickup - Standard, Mobileye FDOT District 7, Administrative, Car FDOT District 7, Administrative, Car, Mobileye HART Large Bus HART Small Bus 120 100 80 60 40 20 0 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 150 100 50 0 Oct 2014 Nov 2014 Dec 2014 Jan 2015 Feb 2015 Mar 2015 HART, Administrative, Bus - Large, HART Maintenance, HART Safety HART, Administrative, Bus - Small, HART Maintenance, HART Safety HART, Administrative, Bus - Large, HART Maintenance, HART Safety, Mobileye HART, Administrative, Bus - Small, HART Maintenance, HART Safety, Mobileye

RESULTS TOTAL MOBILEYE EXCEPTIONS DENSITIES

RESULTS FORWARD COLLISION EXCEPTIONS DENSITIES

SUMMARY AND NEXT STEPS Noticeable change in driver behavior by number of incidents! Reduced number of incidents However, don t notice in statistically significant changes in the vehicles with and without Mobileye Critical to know context of operations Develop performance measures to quantify change due to these devices

CONTACT Krishnan Viswanathan CDM Smith viswanathank@cdmsmith.com 850-386-9500

SIS BOTTLENECK STUDY Chris Edmonston, AICP and Praveen Pasumarthy, P.E., PTOE

INTRODUCTION! Background! Study Purpose! Scope of Study! Study Findings! Next Steps

BACKGROUND! Strategic Intermodal System (SIS) was created in 2003! Designate facilities and services of statewide significance! Serve the mobility needs of citizens, businesses, and visitors! Help Florida to " become a worldwide economic leader, " enhance economic prosperity and competitiveness, " enrich quality of life, and " reflect responsible environmental stewardship.

BACKGROUND

STUDY PURPOSE! Identify bottlenecks on Strategic Intermodal System (SIS)! Use the data available with FDOT! Identify bottlenecks at statewide and district level! Present the results in easy to understand format

SCOPE OF THE STUDY! Activity 1: Data review " Data available with FDOT " Vehicle probe data! Activity 2: Methodology to identify bottlenecks " Research efforts in the field of congestion and bottlenecks " Contemporary practices by DOTs, MPOs, etc. " Develop the methodology! Activity 3: Identify bottlenecks " Analyze the data " Prepare maps/graphics

DATA SOURCES! Existing FDOT Data Sources " Roadway Characteristics Inventory (RCI) " Traffic Monitoring Program # Telemetered Traffic Monitoring Sites (TTMSs) # Portable Traffic Monitoring Sites (PTMSs)! Vehicle Probe Data " INRIX s Vehicle Probe Data Phase I " FHWA s NPMRDS from HERE Phase II

VEHICLE PROBE DATA! Data sources " Traffic sensors # Induction loop sensors # Radar sensors # Toll tag readers " Probe Vehicles! Data is normalized to account for major events, weather conditions, etc.! Data provided in CSV format! Speed data obtained at 5-min intervals! 64 million records (auto); 17 million records (freight)

FHWA S NPMRDS ON NHS

FHWA S NPMRDS ON SIS

COVERAGE OF FHWA S NPMRDS

COVERAGE OF FHWA S NPMRDS Passenger Data Freight Data

COVERAGE OF FHWA S NPMRDS Hourly Daily Monthly 5 14 14 12 12 4 10 10 Observations (in millions) 3 2 Observations (in millions) 8 6 Observations (in millions) 8 6 4 4 1 2 2 0 12 AM 4 AM 8 AM 12 PM 4 PM 8 PM 0 Mon Tue Wed Thu Fri Sat Sun 0 Jul Aug Sept Oct Nov Dec

COVERAGE OF FHWA S NPMRDS Hourly Daily Monthly 5 12 12 Passenger Observations (in millions) 4 3 2 1 Observations (in millions) 10 8 6 4 2 Observations (in millions) 10 8 6 4 2 0 12 AM 4 AM 8 AM 12 PM 4 PM 8 PM 0 Mon Tue Wed Thu Fri Sat Sun 0 Jul Aug Sept Oct Nov Dec 5 12 12 Freight Observations (in millions) 4 3 2 1 Observations (in millions) 10 8 6 4 2 Observations (in millions) 10 8 6 4 2 0 12 AM 4 AM 8 AM 12 PM 4 PM 8 PM 0 Mon Tue Wed Thu Fri Sat Sun 0 Jul Aug Sept Oct Nov Dec

METHODOLOGY TO IDENTIFY BOTTLENECKS! Summary of research efforts in the field of congestion and bottlenecks! Contemporary practices " National performance reports " State DOT practices " MPO/Local government practices! Recommended methodology to identify bottlenecks on SIS network

CONGESTION! FHWA defines congestion as an excess of vehicles on a roadway at a particular time resulting in speeds that are slower - sometimes much slower - than normal or free flow speeds Poor Signal Timing, 5% Bad Weather, 15% Special Events, 5% Bo#lenecks, 40%! Recurring and non-recurring congestion Work Zones, 10% Traffic Incidents, 25%

BOTTLENECK! Bottleneck can be defined as a localized section of highway that experiences reduced speeds and inherent delays due to a recurring operational influence or a nonrecurring impacting event! Recurring and non-recurring bottlenecks! Bottleneck occurs on a subordinate segment of a facility! Bottleneck certainly constitutes congestion, but congestion is often more than a bottleneck

BOTTLENECK

CONTEMPORARY PRACTICES! National Performance Reports " Urban Congestion Report by FHWA " Urban Mobility Report by Texas Transportation Institute " National Traffic Scorecard by INRIX! State DOT Practices " Florida DOT " California DOT " Washington State DOT! MPO/Local Government " Metroplan Orlando " North Florida TPO " Georgia Regional Transportation Authority

RECOMMENDED METHODOLOGY STEP 1: Determine the purpose STEP 2: Develop a plan based on the uses and users STEP 3: Collect and process required data! Primary purpose would be for planning purposes! Bottlenecks should be identified using maps! Resulting product should be easy to understand and interesting to read STEP 4: Calculate performance measures and identify bottleneck & congestion STEP 5: Communicate in a meaningful way

RECOMMENDED METHODOLOGY STEP 1: Determine the purpose STEP 2: Develop a plan based on the uses and users STEP 3: Collect and process required data STEP 4: Calculate performance measures and identify bottleneck & congestion! Bottlenecks identified for auto and truck mode! Bottlenecks identified based on weekday travel patterns! Travel time reliability measures are used! Yearly trends will be reported! Statewide and district-level measures will be reported STEP 5: Communicate in a meaningful way

RECOMMENDED METHODOLOGY STEP 1: Determine the purpose STEP 2: Develop a plan based on the uses and users! Vehicle probe data from FHWA will be used! FDOT data will be used for mapping purposes STEP 3: Collect and process required data STEP 4: Calculate performance measures and identify bottleneck & congestion STEP 5: Communicate in a meaningful way

RECOMMENDED METHODOLOGY STEP 1: Determine the purpose STEP 2: Develop a plan based on the uses and users! Process FHWA NPMRDS data! Identify performance measures! Define and identify bottlenecks STEP 3: Collect and process required data STEP 4: Calculate performance measures and identify bottleneck & congestion STEP 5: Communicate in a meaningful way

RECOMMENDED METHODOLOGY STEP 1: Determine the purpose STEP 2: Develop a plan based on the uses and users STEP 3: Collect and process required data! Display results in easy-to-understand tables and graphics! Present the yearly trends! Present results statewide and by district STEP 4: Calculate performance measures and identify bottleneck & congestion STEP 5: Communicate in a meaningful way

QUANTIFY CONGESTION/BOTTLENECK! NCHRP Report 398: Quantifying Congestion (1997)! Presented methods to measure congestion! Difficult to conceive of a single value to describe congestion! Four components in a congested roadway system " Duration amount of time congestion affects the travel system " Extent number of people or vehicles or geographic area affected by congestion " Intensity severity of the congestion " Reliability variation in the other three elements

QUANTIFY CONGESTION/BOTTLENECK Phase I Methodology Phase II Methodology Duration: Extent: Duration: Extent: Frequency of Congestion None Frequency of Congestion Vehicle Miles Traveled Intensity: Reliability: Intensity: Reliability: None Planning Time Index Vehicle Hours of Delay Planning Time Index

STATISTICAL VALIDATION! Statistical validation of measures is essential! Margin of error " Expresses amount of random sampling error in results " High Margin of error = Low confidence in results " Segments with margin of error greater than 10% are eliminated

STATEWIDE SIS BOTTLENECKS

NEXT STEPS! Phase II of the study is underway to identify the causes of top bottlenecks and suggest corrective actions! Incorporate the bottlenecks in FDOT s Strategic Investment Tool! Update the list of top bottlenecks every two years! Update methodology using advancements in and coverage of vehicle probe data