How autonomous driving could transform road safety
Most crashes are not caused by mechanical failure or bad roads. They are caused by people. This deck argues that a fully autonomous vehicle removes three specific human limitations, and shows the numbers behind each.
Originally written as a research paper for an academic English course; rebuilt here as slides. References on the last page.
The weakest component in the transportation system is the driver
What "autonomous" means here: SAE Level 5
| SAE level | Who drives |
|---|---|
| 0 | No automation; the human does everything |
| 1–2 | Driver assistance; the human supervises constantly |
| 3 | Conditional automation; the human must take over on request |
| 4 | High automation within a defined domain (e.g. Waymo today) |
| 5 | Full automation, all conditions, no human intervention |
Source: SAE International, J3016 (2021).
A machine is not biological, so it cannot get tired
Specific to trucking, and police-reported data are widely believed to understate fatigue across every category of driver.
- Hours-of-service rules and mandatory electronic logging devices have not eliminated the risk. Fatigue is driven by biology, and laws cannot override biology.
- An autonomous vehicle has no body clock and no sleep debt. It runs for as long as the power supply does.
- Over 7.1 million rider-only miles across several cities, Waymo's deployed Level 4 system recorded no fatigue-related failures (Kusano et al., 2024).
It follows the rules, every time
- An automated system operates inside a clear, bounded rule set. It will not speed or run a yellow light to save time, or change lanes without signalling, because legal alignment is designed in.
- In Waymo's deployed fleet, Teoh and Kidd (2017) found no records of speeding, drunk driving or red-light running, behaviours that are routine among humans.
It sees more, farther, and faster than a human eye
- Glare. After oncoming headlights, vision takes about 0.8 s to recover for younger drivers and up to 2.1 s for older ones (Reuten et al., 2018). That gap contributes to night-time crashes.
- Field of view. Combined horizontal vision approaches 200°, but only about 60° is sharp. Everything else needs a head turn or a mirror.
Waymo, 2024. The industry is split between multi-sensor fusion (Waymo) and vision-only (Tesla); which prevails is unsettled, but both far exceed human perception.
The advantages are already measurable
An 85% reduction in injury-reported crashes, on a Level 4 system that still has limits a Level 5 system would not.
Kusano, Scanlon, Chen, McMurry, Chen, Gode and Victor (2024), Traffic Injury Prevention.
Three limitations removed, three documented causes of death addressed
Sources
· Federal Motor Carrier Safety Administration (2006). Report to Congress on the Large Truck Crash Causation Study. US DOT.
· Kusano, K. D., Scanlon, J. M., Chen, Y.-H., McMurry, T. L., Chen, R., Gode, T., & Victor, T. (2024). Comparison of Waymo rider-only crash data to human benchmarks at 7.1 million miles. Traffic Injury Prevention, 25(sup1), S66–S77.
· National Highway Traffic Safety Administration (2024). Overview of motor vehicle traffic crashes in 2022 (DOT HS 813 560).
· Reuten, A. J. C., et al. (2018). Headlight glare exposure and recovery in younger and older drivers.
· SAE International (2021). Taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles (J3016_202104).
· Singh, S. (2015). Critical reasons for crashes investigated in the National Motor Vehicle Crash Causation Survey (DOT HS 812 115). NHTSA.
· Teoh, E. R., & Kidd, D. G. (2017). Rage against the machine? Google's self-driving cars versus human drivers. Journal of Safety Research, 63, 57–60.
· Waymo (2024). Meet the 6th-generation Waymo Driver. Waymo blog.