Street Safety - Techie Topics

 

When a speeding motorist is pulled over by a police officer and he asks, "Do you know how fast you were going?", most people say "not exactly", and probably often this is true, especially on city streets where cruise control is not much used and the stroad feel is more highway than street.
We say individuals are responsible for managing their speed, yet we give them more tools to evade laws than to follow them, and then we're somehow surprised when the behavior is less than exemplary.

• Why are cars built to go over twice as fast as any highway speed limit in the country?
• Why do cars permit setting cruise control in cities, and at speeds over the speed limit on highways?
• Why do uninsured cars operate at all?
• Why do cars, which know exactly where they are on the city streets, not at all control speed on those streets to match posted limits?
• Why can cars with hacked vehicle computers and inoperable safety systemsbe operated on the street?
• How can we have meaningful crash analysis without tracking data from the cars available to investigators?
• Why do cars detect other cars but generally not pedestrians?
• Why do cars have cushioning for internal occupants but steel for external impacts?

These are mostly rhetorical questions, as the answer for all is that the pattern for individual liability, not vehicular design, and certainly not manufacturer liability, was set long before "smart" car technology came along, and has evolved further to favor car manufacturers. Clearly, vehicle manufacturers have done a pretty good job of eschewing responsibility for the operation of cars and for deaths inside and outside the car.   Our government also shares some culpability, with heavy regulations for safety of occupants but not much for crash causation and external hazards. And, of course, our libertarian thinking tends to extend to our cars, with a strong desire by drivers to have control of their cars and of heir  data with a strong bias to not have cars self-incriminate their drivers.

All of this is an area where NTSB could help out, as their EU peers already do. A few things that could quite readily be accomplished by federal mandate:

• Add a "black box" feature available to crash investigators and insurance for crash recreation. Note that the EU already is doing this: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwjPyNfX2bDzAhXfQzABHS-zDs0QFnoECAQQAw&url=https%3A%2F%2Fwww.jdsupra.com%2Flegalnews%2Fblack-boxes-in-automobiles-european-5035475%2F&usg=AOvVaw0aUE7cNU6zRfjH-KuVWaXv
• Limit speeds on city streets to the posted limits -- no reason for speeding or passing in school zones, and no "10 over" buffer that enforcement generally tolerates. The EU is already doing this for 2022 vehicles: https://www.theguardian.com/world/2019/mar/27/all-new-uk-cars-to-have-speed-limiters-by-2022-under-eu-plans
• Incorporate pedestrian sensors and automated braking systems, with added sensors for blind-spot sensors. EU regulators are adding some such sensors already for 2022.
• Support monitoring by insurance. Sure, accelerating to beyond the speed limit to pass may be "safe and appropriate", but such situations should be brief and infrequent. Patterns of speeding should be noted and ideally self-limited by the car. To begin with, these can be optional, but as crash data accumulates the cost premium to NOT be monitored should naturally increase to match experienced risk. Some insurance companies support such monitoring already, but the cost savings for using what is considered to be an invasive technology is not large. Similarly, insurance reductions for other safety systems aren't large either. In some cases, this is due to lack of historical data and validation studies, but also to relatively low liability for crash injuries and deaths.
• Add integrated cameras, front and back. This is an obvious point of value, and easy to implement. Simply put, people lie. Like black-box vehicle data, video should be part of crash data recording.
• Integrate universal "OnStar"-style crash reporting, where the vehicle alerts the manufacturer and local safety agencies of any detected crash. Hit-and-run evasion of responsibility would plummet with automated crash monitoring
  

The above items are relatively straightforward, and should be "go do" items for our regulators and, if necessary, Congress.  As we increase the value of human life, the resulting liability for crashes can help justify further technological development.  Self-driving cars will need all of this technology, and instead of adding new risks we should first use such tech to reduce current risks, especially to those not in the car.

• Night vision enhancements, highlighting obscured individuals and animals.
• Automatic detection of rain, fog, and lighting conditions where speeds should be reduced, but where habit and complacency add risk today.
• Add firmware signing and crytographically secure boot and software attestation to the car control software, such that mods and hacks to eliminate manufacturer emissions controls, data collections, and monitoring will be more difficult.
• Close the loop between car and insurance company, such that the car "knows" its insurance info and insurance company "sees" the car operation, and an uninsured car simply doesn't operate beyond a low-speed limp-home mode.
  
• Authenticate users with vehicles, such that new drivers get extra supervision and feedback on driving patterns.
• Require yearly inspections of vehicle controls and emissions, exhaust noise level, window tints, safety equipment, monitoring systems, tires and lights, etc.
• Create options for cities to ring-fence vehicles by GPS, such that some types of vehicles are not permitted in city centers without permits or additional controls. For example, no combustion engine vehicles downtown, or no trucks without appropriate levels of safety equipment, or large trucks without special permits.

Many will say these features are invasions of privacy, and perhaps they are, but driving is a privilege that is massively abused, and our goal should be to make driving in cities utilitarian and not exciting -- exactly the opposite of what car advertising provides. Roaring engines? Exhilarating acceleration? Smoky burnouts? Intimidatingly tall grills and lifted bodies? Why would we want any of these things on our city streets, especially near kids on foot or riding bicycles?

It's past time for the NTSB to take the lead in these regulations, and at least keep pace with what the EU is doing. Ideally, we would be leaders in the safety realm, using every crash as an opportunity to learn and thereby improve safety. We need to swing the pendulum back from libertarian-style individualism to broader collective responsibility if we are to significantly reduce the 40K deaths and millions of injuries per year, especially those of more vulnerable road users.