School of something FACULTY OF OTHER Help or hindrance? Demand implications of vehicle automation Zia Wadud Associate Professor in Transport and Energy Institute for Transport Studies & Centre for Integrated Energy Research
Objective To identify the directions of travel impacts of highly/fully automated vehicles (with some quantification) To identify the key areas that require attention
Levels of automation No automation Eyes On Driver assistance Eyes on Partial automation Eyes On Conditional automation High automation Eyes Off Mind Off Full automation Mind Off
The ripple effect New users Existing users Trip distance Trip numbers Mode choice Own vs. mobility on demand Wadud & Anable 2016
New users 11.6 million disabled people in the UK 6.5 million mobility-impaired Immense wellbeing benefits Younger generation parental escort vs. driverless escort? USA: elderly 2-10% increase in demand Wadud et al. 2016
Existing users Parking & empty running Trade-off: Public transport vs. private car VTTS will certainly be lower, but how much lower? Trip distances, trip rates Role of time use and VTTS crucial Time, not material goods, raises happiness -BBC News this morning
Existing users Time use How do people intend to use time in automated cars? How do people use time in cars now? Is there a correlation between perceived usefulness of travel time and intended use of automated cars? What is the effect of motion sickness? The commute felt like it took half the time -Tesla autopilot user
Existing users Time use results: Current vs. intended time use Activity on which most time is or will be spent Revealed ranking for current car passengers Ranking of stated intention in future FAVs Still watching roadway - - Working/ studying 2 1 Window gazing/ people watching 4 4 Thinking/ planning 1 3 Phone calls/ messaging 9 10 Online social media 5 6 Reading for leisure 10 7 Emailing/ browsing internet 7 8 Eating/ drinking 11 12 Sleeping/ snoozing 7 9 Listening to music/ radio 6 5 Watching video/ playing games 12 11 Talking to other passengers 2 2 Rank correlation 0.92 Wadud & Huda 2017
Existing users Time use results: Perceived usefulness of travel time & interest to use automated vehicles Commute 140 120 100 Leisure 80 60 40 20 0 Less than 10% time more useful 10-30% time more useful 30-50% time more useful 50%-75% time more useful More than 75% time more useful 180 160 140 120 100 Very interested Moderately interested Slightly interested Not interested 80 Supports multitasking & mode choice literature Wadud & Huda 2017 60 40 20 0 Less than 10% time more useful 10-30% time more useful 30-50% time more useful 50%-75% time more useful More than 75% time more useful Very interested Moderately interested Slightly interested Not interested
Own vs. MoD Automated vs. manual vs. automated taxi operations: costs 25,000 20,000 15,000 VED, breakdown Tyre, maintenance, parking Insurance Depreciation Cost of capital Fuel Wasted travel time 140,000 120,000 100,000 80,000 License, overhead Tyres, maintenance Insurance Depreciation Cost of capital Fuel Wages 10,000 60,000 40,000 5,000 20,000 0 Current Auto Current Auto Current Auto Current Auto Current Auto Current Auto 0 Current Auto Current Auto Cu Lowest quintile 2nd quintile 3rd quintile 4th quintile Highest quintile 99th percentile Taxi 7.5 Tonne Rigid truck 18 Wadud 2017
Own vs. MoD MoD: Marginal cost pricing should curb demand in theory Self-selected bunch Empty running Public transport to on-demand-services? Vicious circle VMT won t fall unless ridesharing ; Evidence of some sharing but who uses MoD? Some capacity benefits through rightsizing MoD. But induced traffic Cars parked 95% of the time; 1 car club car removes 9 cars on street; Does it matter?
Travel demand Mechanisms Impacts Automation level Distances Modal shift Trip number New user groups Mobility on demand Empty running Smaller impact at low levels of automation Step change at high levels of automation But demand will almost certainly increase USA: up to 60% increase in demand, range 5% (low levels)-60% (full automation) Wadud et al. 2016
Index Travel demand Two important hypothesis challenged Marchetti constant/ Zahavi Travel time budget 200 Car VKT/capita 190 BBE prediction 180 WB1 prediction Peak car 170 1 hr/day commute 160 150 140 130 120 110 100 1980 1985 1990 1995 2000 2005 2010 2015 Year
Business/freight demand Total cost of ownership analysis (UK) for trucks 140,000 120,000 100,000 80,000 License, overhead Tyres, maintenance Insurance Depreciation Cost of capital Fuel Wages 60,000 40,000 20,000 0 Current Auto Current Auto Current Auto Current Auto Taxi 7.5 Tonne Rigid truck 18 Tonne Rigid truck 38 Tonne Trailer-truck Same conclusion: VMT by trucks, vans, lorries goes up Wadud 2017
Uncertainties VTTS will certainly fall, but by how much? VTTS vs. value of reliability? Own vs. ridehail vs. rideshare: Not either/or all will coexist What is the equilibrium share? Who & where from switch occurs? Overall VMT will almost certainly go up, absent any policies Although some directions uncertain trips escorting children? How much will it go up? Urban or intercity? Which types? Nearly all studies use stated preference/intention Any evidence from revealed behaviour?
School of something FACULTY OF OTHER Thank you The use and usefulness of travel time in fully automated vehicles, 2017 (under review) Fully automated vehicles: A cost of ownership analysis to inform early adoption, 2017 Help or Hindrance? The travel, energy and carbon impacts of highly automated vehicles, 2016
Energy consumption impacts (USA) Platooning Eco-driving Congestion mitigation De-emphasized performance Improved crash avoidance Vehicle right-sizing Higher highway speeds Increased features Travel cost reduction New user groups Changed mobility services Infrastructure footprint* -60% -50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50% 60% % changes in energy consumption due to vehicle automation