Connected / Automated Vehicles (CAV) Traffic Forecasting PerspectiveJosiah Banet, P.E.

July 24, 2018

Overview

• Background for Modeling CAVs

• CAV Transportation Impacts

• Transportation Supply Side

• Transportation Demand Side

• Modeling Approach

• Future Applications

2

Background for Modeling CAVs

• Part of future transportation network

• Not about if, but when

• Potential disruptor

• Widespread interest and opinions

• New questions from project

stakeholders

• CAV benefits

• Traffic modeling looked to for answers

3

Background for Modeling CAVs

• Traffic modeling is data dependent

• Lack of actual data for CAVs

• New modeling paradigm

• Need consistent methodology

• Long-term projects

• Another reference point

4

CAV Transportation Impacts

• Hot topic in transportation

• Substantial research

• Transportation Supply

• Available network capacity

• Transportation Demand

• Number of trips on network

• No consensus

• Magnitude

• Effect

• Timing

5

Transportation Supply• Increased roadway capacity

• Shorter gaps between vehicles• Mixed traffic conditions• Perfect driver

• Vehicle platooning• Reduced right-of-way needs

6

Transportation Demand

• Land use• Where people live and work

• Trip production• Number and types of trips

• Trip distribution• Trip lengths

• Mode choice• New modes

• Assignment • Route choice

7

Transportation Demand• Mobility independence

• Induced trips by seniors, elderly, children, etc.

• Zero occupancy vehicles• Auto ownership• Mode shifts (ride-sharing)• Longer commutes/trips

8

Fleet Adoption Timeline

• Market Penetration is overriding input• Determines other assumptions

• Estimates vary• Year and magnitude

• Factors• Vehicle capabilities• Price• Retrofitting existing vehicles

• Adoption rate will drive design• Non-uniform impacts

• Transportation supply• Transportation demand

9

CAV Market Penetration

Modeling Approach

• Network flexibility• Model variables

• Transportation supply • Transportation demand

• Control Inputs • Highway network• Land use• Value of time• Trip lengths

• Sensitivity analysis

10

Modeling Approach: Transportation Supply

11

-

10.0

20.0

30.0

40.0

50.0

60.0

70.0

100 500 900 1300 1700 2100 2500 2900 3300 3700 4100

Spee

d (m

ph)

Volume per lane

Congested Speed by CAV Percentage

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

BPR Volume Delay Curve

• Increase capacity with higher CAV adoption

• Impact will vary by facility type• Dependent on the percentage of

CAVs on roadway link• Focus on shorter gaps• Facility preference

Modeling Approach: Transportation Demand

• CAV ownership and use• Initial CAV adoption in certain areas

• Urban areas• Affluent areas• Similar to other technology adoption trends

• Followed by widespread adoption• Safety benefits• Transportation benefits

• New trips

12

Modeling Approach: Central Florida CAV TAZ Inputs

13

Population Density

LegendCFRPM6TAZPopulation Density

< 1k

1k - 2k

2k - 4k

4k - 8k

> 8k

Traffic Analysis Zone (TAZ)Population Density (Persons/Sq Mile)

Property Value

LegendCFRPM6TAZHouse Price

<75K

75K - 150K

150K - 225K

225K - 450K

>450K

Traffic Analysis Zone (TAZ)Average Property Value ($)

Modeling Approach: Central Florida CAV Trips

14

50% CAV Regional Adoption

LegendTAZCAV Percentage

< 10%

10% - 20%

20% - 30%

30% - 40%

40% - 50%

> 50%

Traffic Analysis Zone (TAZ)CAV Percentage

20% CAV Regional Adoption

LegendTAZCAV Percentage

< 10%

10% - 20%

20% - 30%

30% - 40%

40% - 50%

> 50%

Traffic Analysis Zone (TAZ)CAV Percentage

Modeling Approach: 2050 Statewide Network

15

2050 Statewide Network Comparison

Non-CAV 50% CAV Adoption Change

Vehicle Miles Traveled (VMT) 687,245,676 717,845,808 +4.5%

Vehicle Hours Traveled (VHT) 16,838,945 16,130,688 -4.2%

Average Network Speed (mph) 40.81 44.50 +9.0%

Average Network Volume/Capacity 0.54 0.42 -22.2%

• Non-Uniform CAV ownership• Non-Uniform distribution of CAV trips• Dynamic roadway impacts• Overall network travel increases while

delays decrease

2050 50% Statewide CAV Adoption

LegendHWYLOAD_CAV_1CAV Percentage

<30%

30% - 40%

40% - 50%

50% - 60%

60% - 70%

>70%

LegendNetwork

CAV Percentage

Modeling Approach: 2050 Network Congestion

µ

2050 50% Statewide CAV Adoption

LegendNetworkVC Ratio

<0.6

0.6 - 0.7

0.7 - 0.8

0.8 - 0.9

> 0.9

LegendNetwork

Volume / Capacity Ratio

16

2050 0% CAV Statewide Adoption

LegendNetworkVC Ratio

<0.6

0.6 - 0.7

0.7 - 0.8

0.8 - 0.9

> 0.9

LegendNetwork

Volume / Capacity Ratio

Future Applications

• Analyze model results• Compare with Non-CAV results• Identify trends and areas of impact• Long-term project traffic evaluation

• Consider other variables• Trip lengths• Land use changes• Parking

• Other model years• Other model types

17

Questions?

18