When Daunte Wright was fatally shot by an officer during a traffic stop near Minneapolis this week, he joined a long list of people who have died at the hands of the police after being pulled over for a traffic violation. Traffic stops should not be harrowing or dangerous experiences, but too often they are for people of color.

One way to address this problem is to reduce the number of encounters that drivers have with police officers. At the same time, any responsible reform must account for the fact that accidents involving motor vehicles are a leading cause of death for Americans under the age of 54. Road safety is itself a serious problem, one that requires laws and regulations that must be enforced.

How can we reduce traffic stops without undermining public safety? The solution is to decrease our reliance on human enforcement. Having police officers implement traffic laws is not the only way to promote road safety. Indeed, the evidence suggests that it is not even the optimal way to do so.

Automated speed cameras and red-light cameras, for example, have proved to be effective in decreasing traffic accidents, injuries and fatalities, precisely because they're more consistent than human oversight. They also don't selectively — or discriminatorily — choose to pull over violators. Automating citations for speeding, a major cause of accidents, could significantly reduce police encounters. In New York state, for example, speeding accounts for nearly 20% of all traffic citations, according to recent data.

Similar technology could be used to check for expired licenses and registrations — common infractions that disproportionately ensnare the poor and people of color. (Expired registration was the reason the police gave for stopping Wright.) Many police departments already use automated license-plate readers. Instead of having human officers issue citations for this violation, license-plate readers could be used to send notice by mail that a license or registration has expired or is about to.

As with any technological solution, we would need to be mindful of individuals' privacy rights. Without sufficient protections in place, the police could cull a great deal of data from traffic cameras and license-plate scanners and misuse that information.

Decreasing our reliance on human enforcement of traffic laws does not always mean switching to automated enforcement. Better design of streets and highways can protect lives, too. The legal scholar James Willard Hurst once remarked that the people who figured out how to draw a line down the middle of the road did more to enforce the keep-to-the-right rule than anything the police could do. Raised medians can help keep bikers safe from cars (and even beautify the streets), which regulations on separate lanes can't do. Let's think about speed bumps as well as speed limits.

Of course, some human enforcement of traffic laws will always be necessary. But traffic police do not need to be weaponized officers trained to approach stops with a heightened expectation of conflict; unarmed traffic monitors could be used instead. The police would still pursue criminal violations, and they could provide assistance when requested by traffic monitors. But an armed response would be a last resort, not routine.

This makes statistical sense: Available evidence shows that a negligible percentage of routine traffic stops result in serious injury or death to officers. That number is just 0.31% in Florida, where a bill is pending to create a public safety department with traffic monitors who would enforce moving infractions. Transferring enforcement of civil traffic laws to a nonpolice agency can decrease police encounters on the road and still protect drivers and pedestrians.

These ideas are not entirely new. The father of modern policing, August Vollmer, who was the police chief in Berkeley, Calif., from 1909 to 1932, argued that traffic enforcement distracted police from their main job of fighting crime. His insight was soon forgotten, however, as American society became a car society and as the getaway car often accompanied the commission of crime. Law enforcement came to depend on investigatory stops — that is, traffic stops conducted primarily for the purpose of looking into criminal activity.

The strategy of investigatory stops has not affected crime rates, according to a recent study by the Policing Project at New York University's law school. But it has served to alienate and harm those unjustifiably targeted for inspection. Vollmer was right that traffic duty is a distraction from the goal of solving crime. What he didn't know was that — in terrible, tragic ways — police enforcement of traffic laws would also come to undermine the primary purpose of traffic stops: public safety.

Sarah A. Seo is a professor at Columbia Law School and the author of "Policing the Open Road: How Cars Transformed American Freedom." She wrote this article for the New York Times.