Accounting for AV/CV in Long- Range Plans Using Current Travel - PowerPoint PPT Presentation

Accounting for AV/CV in Long- Range Plans Using Current Travel Demand Models Presented to 2016 TRB Tools of The Trade Conference

Background • 25 MPO’s – 13 TMAs – 12 Non-TMAs

Background Traffic – Three urban areas in top 12 most congested urban areas (TTI Report) • Houston, Dallas, Austin – Austin has worst congested roadway in Texas

Background Models – 1 ABM model – 24 Trip-based models • 4 study areas with full mode choice • Handful with mode shares • Majority are 3- step with direct vehicle generation

Background • Forecasting AV/CV demand – 4 th task in larger research supported by TxDOT – How does one measure the potential impacts across the state? • Consistent guidance, approaches and measurable outcomes desired by TxDOT

Assumptions • 100% vehicle mix – Fully autonomous and connected – Consistent with NHTSA Level 4 definition • Current household auto ownership levels maintained – Relinquish navigation, or – Participate in shared-rides (albeit limited)

Assumptions • Vision of greater ride-sharing – Carpooling in tours – “Robo - Taxis” • Difficult to predict acceptance or system • Therefore: – Shared-ride splits are held constant, or – Proportionally adjusted based on existing forecasted mode shares

Assumptions • VMT of unoccupied “robo - taxis” not accounted for in study • All sectors in urban area treated equally – Restrict travel within downtown, for example • Existing external splits held constant

Identifying A Study Area • Enumerating demand or system changes, although possible, magnitude of changes may be limited in a majority of MPOs in the State of Texas – Limited appreciable system-wide congestion – Limited transit ridership (no mode choice model) – Narrow peak periods and/or spot congestion

Study Area • Austin, Texas (CAMPO) – Six-county study area – System-wide congestion – Most-congested roadway in Texas – Extreme peaks – Transit component

Study Area • Austin, Texas (CAMPO) – Population growth 64% (2010 to 2040) 2040 2010 Williamson Travis Hays Caldwell Burnet Bastrop 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 1,600,000 Population Source: Texas State Data Center

CAMPO TDM • 4-Step travel model Develop Input Files – Similar trip generation and Initialization distribution models to TxDOT Trip Generation (Texas Package) Trip Distribution – Mode choice model Mode Choice • Nested-logit model (auto, Feedback transit, non-motorized) Trip Tables – 15-trip purposes Trip Assignment – Generalized-cost assignment Model Reports Source: CAMPO TDM Validation Report

Identifying Scenarios • Balance between reasonable assumptions and optimistic enthusiasm – Fleet turnover – Shared rides – Greater mobility for different cohorts (e.g., age & disabled) .... • Uncertainty • Arguments and counter-arguments

Identifying Scenarios • “Typical” items that could be given consideration Land Use Utility of Freight Travel External Costs Travel TDM & AV/CV Network Trip Length Capacity Time Choices Trip Generation Mode Routes Choice

Identifying Scenarios • Unintended consequences & outcomes Education Land Use Utility of Freight Primary & Travel Secondary Retail Workplace Scope and External Location Costs Travel Location TDM & AV/CV Land Network Trip Length Capacity Household Freight Location Distribution Use Time Choices Trip Generation Mode Routes Choice

Scenarios Scenarios “Base” S1 S2 S3 S4 S5 S6       Limited Limited Limited Limited Limited Limited increase in increase in increase in increase in increase in increase in EXPWY EXPWY EXPWY EXPWY EXPWY EXPWY and FRWY and FRWY and FRWY and FRWY and FRWY and FRWY capacity capacity capacity capacity capacity capacity      Increase Increase Increase Increase Increase per hour per hour per hour per hour per hour per lane per lane per lane per lane per lane capacity of capacity of capacity of capacity of capacity of FRWY FRWY FRWY FRWY FRWY 2040 links links links links links MTP     Increase Increase Increase Increase arterial arterial arterial arterial Forecast capacity capacity capacity capacity by 10% by 10% by 10% by 10%    Proportion Proportion Proportion ally move ally move ally move transit transit transit trips to trips to trips to SOV and SOV only HOV trip HOV (2 & trip table tables. 3+) trip tables Sequential & cumulative results

Scenario Assumptions • Study limited to system & choice • Model inputs held constant: Choice – Demographics • Household and workplace location TDM – Trip rates – External forecasts Demand System – Trip lengths • Observed data non-existent – Imposing assumptions

Scenario Results • AM period results only – VMT – Speeds – Travel Time – Delay – VMT per person – Average trip lengths in minutes and miles – Modes

Scenario VMT Results Total AM VMT VMT Growth (Scenario to Baseline) Scenarios 8.79% 18.5 9.00% 7.85% 7.92% Base S1 S2 S3 S4 S5 S6 AON 7.50% 7.14% 6.86% 16,795,034 17,187,458 17,947,172 17,993,762 18,112,750 18,124,662 18,055,190 8.00% 18,270,971 18.0 7.00% AM VMT (MILLIONS) 6.00% 17.5 VMT Growth 5.00% 4.00% 17.0 3.00% 2.34% 2.00% 16.5 1.00% 16.0 0.00% Base S1 S2 S3 S4 S5 S6 AON* Scenarios

Scenario VMT Results 0.00 - 0.85 0.85 - 1.00 1.00 - 1.15 1.15 and above Proportion of Uncongested Travel 18.0 100.00% 93.02% 92.87% 92.84% 92.94% 91.10% 90.00% 16.0 81.16% 80.00% 74.85% 14.0 Proportion of Uncongested Travel 70.00% AM VMT by V/C Ratio (Millions) 12.0 60.00% 10.0 50.65% 50.00% 8.0 40.00% 6.0 30.00% 4.0 20.00% 2.0 10.00% 0.0 0.00% Base S1 S2 S3 S4 S5 S6 AON* Scenarios

Scenario VMT Results 2040 MTP Results “Base” 2040 Scenario 3

VMT (MILLIONS) -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 0.0 - 7.5 7.5 - 12.5 Scenario Speed Results 12.5 - 17.5 17.5 - 22.5 22.5 - 27.5 S1 S2 27.5 - 32.5 S3 Speed Bins S4 32.5 - 37.5 S5 S6 37.5 - 42.5 42.5 - 47.5 47.5 - 52.5 52.5 - 57.5 57.5 - 62.5 > 62.5

Scenario Speed & VMT Results <= 22.5 mph 22.5 - 42.5 42.5 - 62.5 > 62.5 4.0 100.00% 80.00% 3.0 60.00% Changes in VMT (MILLIONS) 2.0 Percentage Change in VMT 40.00% 1.0 20.00% 0.0 0.00% -1.0 -20.00% -2.0 -40.00% -3.0 -60.00% S1 S2 S3 S4 S5 S6 Scenarios

Scenario Delay Results VHT_AM / VHT_FF_AM Percentage Change 2.5 0.00% -5.74% -10.00% -12.92% 2 -21.53% -21.53% -22.01% -22.01% -20.00% Percent Change to Base Travel Time Ratio 1.5 -30.00% 1 -40.00% 0.5 -52.15% -50.00% 0 -60.00% Base S1 S2 S3 S4 S5 S6 AON Scenarios

Scenario per Person VMT & Delay Results Change in VMT/Person % Change (Delay per Person) 15.00% 8.79% 10.00% 7.85% 7.92% 7.50% 7.14% 6.86% 5.00% 2.34% 0.00% Percentage Change -5.00% -10.00% -15.00% -20.00% -25.00% -30.00% -35.00% Base S1 S2 S3 S4 S5 S6 Scenarios

Scenario Avg. Trip Length Results Congested ATL (Minutes) Change in ATL (Scenario to Base) 30.00 0.00% -5.18% 25.00 -5.00% Avg. Trip Length (Minutes) 20.00 -10.39% Percent Change -10.00% 15.00 -15.00% 10.00 -18.77% -19.20% -19.38% -20.05% -20.00% 5.00 0.00 -25.00% Base S1 S2 S3 S4 S5 S6 Scenarios

Scenario Avg. Trip Length Results Total AM Trips ATL (Miles) 1.64 11.40 11.19 11.14 11.15 11.20 1.63 11.11 11.05 11.00 Average Trip Length (MILES) 1.62 AM Trips (MILLIONS) 10.80 10.68 1.61 10.60 10.50 1.60 10.40 1.59 10.20 1.58 10.00 Base S1 S2 S3 S4 S5 S6 Scenarios

General Scenario Results Metric Trend Vehicle Miles of Travel (VMT)  Region  Per Person Travel Time  Travel in Uncongested Conditions  Travel in Congested Conditions Congested Weighted Speeds Travel Time Delay Average Trip Length  Minutes  Miles Mode Shares (Transit)

Where Are We Know? • Limited acceptance of placing AV/CV scenario in current plans – Curiosity – Not yet tangible • Leadership and guidance needed to develop consistent approaches and metrics

Special Thanks • Wade Odell (TxDOT Project Manager) • Hao Pang (TTI) • Tom Williams (TTI)

Questions?