Survey of Studies on Life-Saving Potential of Driverless Cars

by on June 30, 2017 · 0 comments

Whatever you want to call them–autonomous vehicles, driverless cars, automated systems, unmanned systems, connected cars, piloteless vehicles, etc.–the life-saving potential of this new class of technologies has been shown to be potentially enormous. I’ve spent a lot of time researching and writing about these issues, and I have yet to see any study forecast the opposite (i.e., a net loss of lives due to these technologies.) While the estimated life savings vary, the numbers are uniformly positive across the board, and not just in terms of lives saved, but also for reductions in other injuries, property damage, and aggregate social costs associated with vehicular accidents more generally.

To highlight these important and consistent findings, I asked my research assistant Melody Calkins to help me compile a list of recent studies on this issue and summarize the key takeaways of each one regarding at least the potential for lives saved. The studies and findings are listed below in reverse chronological order of publication. I may try to add to this over time, so please feel free to shoot me suggested updates as they become available.

Needless to say, these findings would hopefully have some bearing on public policy toward these technologies. Namely, we should be taking steps to accelerate this transition and removing roadblocks to the driverless car revolution because we could be talking about the biggest public health success story of our lifetime if we get policy right here. Every day matters because each day we delay this transition is another day during which 90 people die in car crashes and more than 6,500 will be injured. And sadly, those numbers are going up, not down. According to the National Highway Traffic Safety Administration (NHTSA), auto crashes and the roadway death toll is climbing for the first time in decades. Meanwhile, the agency estimated that 94 percent of all crashes are attributable to human error. We have the potential to do something about this tragedy, but we have to get public policy right. Delay is not an option.

Accelerating the Future: The Economic Impact of the Emerging Passenger Economy (June 2017)

an Intel Report

  • p. 23: “If we conservatively assume that just 5 percent of these accidents are avoided in the decade from 2035 to 2045 due to pilotless vehicles, 585,000 lives will be saved during that time.”

Implications of connected and Automated vehicles on the Safety and Operations of Roadway Networks: A Final Report (Oct 2016)

By The University of Texas at Austin Center for Transportation Research

Chapter 4, Safety Benefits of CAVs

See Table 4.7,4.8, 4.9 (p.95-97) Annual economic cost and functional-years lost savings estimates from safety benefits of CAV technologies

  • p. 78: The most recently-available U.S. crash database (the 2013 National Automotive Sampling System (NASS) General Estimates System (GES) was used, and results suggest that advanced CAV technologies may reduce… functional human-years lost by nearly 2 million (per year, assuming a market penetration rate of 100%)
  • p. 80: Lane Departure Warning (LDW) systems can reduce 47% of all lane-departure-related crashes, corresponding to 85,000 crashes annually
  • p. 80: Backing-crash countermeasures (like backup collision intervention via automated braking) could prevent almost 65,000 backup crashes a year.
  • p. 80: With an assumption of 100% deployment and 100% device availability (for Road departure crash warning (RDCW) technology), an annually reduction of 9,400 to 74,800 U.S. road-departure crashes was predicted.
  • p. 81: V2V systems, such as FCW, blind spot warning (BSW), and lane change warning (LCW), can serve as primary crash countermeasures, reducing U.S. light-duty vehicle-involved crashes by 76%. They further estimated that V2I systems, such as curve speed warning (CSW), red light violation warning system (RLVW), and stop sign violation warning (SSVW), if deployed anywhere they could be useful, could address 25% of all light-duty-vehicle crashes in the U.S. 

Automated Vehicle Crash Rate Comparison using Naturalistic Data (Jan. 2016)

Commissioned by Google, Performed by the Virginia Tech Transportation Institute (Data adjusts for unreported crashes)

  • Estimated crash rates for the Self-Driving Car Project were lower for all three crash levels… Additionally, the rate of less-severe crashes (Level 3) for the Self-Driving Car was lower at a statistically significant level (39)
  • See Table 10 p.41 “Current data suggest that self-driving cars may have low rates of more-severe crashes (Level 1 and Level 2 crashes) when compared to national rates or to rates from naturalistic data sets.”
  • “The data also suggest that less-severe events (Level 3 crashes) may happen at a significantly lower rate for self-driving cars… none of the vehicles operating in autonomous mode were deemed at fault” (p.41)

The Future of Motor Insurance: How Car Connectivity and ADAS are Impacting the Market (2016)

HERE and Swiss Re

  • See p.15, Figure 9: Accident Reduction Rate by Selected Features
  • Advanced ADAS (highway pilot) would reduce accidents on motorways by 45.4% and on other roads by 27.5%
  • Sophisticated ADAS (lane keeping assistant, emergency braking assistant, night vision) would reduce accidents on motorways by 25.7% and on other roads by 27.5%

A Preliminary Analysis of Real-World Crashes Involving Self Driving Vehicles (Oct. 2015)

University of Michigan’s Transportation Research Institute

  • p. 14: The most common outcome of crashes for both vehicle types was property damage only, but self-driving vehicles had this outcome 10% more often than conventional vehicles. Consequently, self-driving vehicles experienced injury-related crashes 10% less often than conventional vehicles. The overall severity of crashes involving self-driving vehicles was also lower than for conventional vehicles.
  • p. 18: Four main findings:
  1. The current best estimate is that self-driving vehicles have a higher crash rate per million miles traveled than conventional vehicles, and similar patterns were evident for injuries per million miles traveled and for injuries per crash.
  2. The corresponding 95% confidence intervals overlap. Therefore, we currently cannot rule out, with a reasonable level of confidence, the possibility that the actual rates for self-driving vehicles are lower than for conventional vehicles.      
  3. Self-driving vehicles were not at fault in any crashes they were involved in.
  4. The overall severity of crash-related injuries involving self-driving vehicles has been lower than for conventional vehicles.
  • Limitations of the study (stating that crash rates for self-driving vehicles are higher than conventional vehicles) are corrected for in the more recent 2016 Google Study (see above), to show that actually self-driving vehicles crash less.

Ten Ways Autonomous Driving Could Redefine the Automotive World (June 2015)

McKinsey Report

  • Suggests that advanced ADAS and AVs could reduce accidents by up to 90%

Connected and Autonomous Vehicles: The UK Economic Opportunity (Mar 2015)


  • p.2 & p.12: By 2030, connected and autonomous vehicles could save over 2,500 lives and prevent more than 25,000 serious accidents in the UK.

Preparing a Nation for Autonomous Vehicles (Oct. 2013)

Eno Center for Transportation

  • p. 8, Table 2: Estimates of Annual Economic Benefits from AVs in the United States
  • 10% market-penetration would mean 1,100 lives saved; 50% would be 9,600 lives; 90% would be 21,700 lives


Previous post: