a

Facebook

Twitter

Copyright 2018 Hyland + Padilla, PLLC


All Rights Reserved | Privacy Policy | Sitemap
Site by Digimatiq Marketing, Inc

8:30 AM - 5:30 PM

Open Mon. - Fri. With 24/7 Live Answering

919.419.9933

Call Us For Free Consultation

Facebook

Twitter

Search
Menu
 

The Danger of Self – Driving Cars for Pedestrians

The Danger of Self – Driving Cars for Pedestrians

raleigh-personal-injury-advice
a

By Raleigh + Durham Worker’s Comp & Personal Injury Attorneys at Hyland + Padilla, PLLC


raleigh-personal-injury-advice

Many studies predict that self-driving cars will bring a huge reduction in traffic fatalities, but we haven’t reached that future yet. Autonomous vehicles (AV) might help diminish accidents vis-à-vis other vehicles, but there’s evidence that they may pose notable risks to pedestrians.

 

Pedestrians tend to behave more unpredictably than vehicles. They might dart out into the street, or meander erratically in a crosswalk while gazing at their cell phone. Human drivers can detect and anticipate such human actions and adjust our speed accordingly. But it’s challenging to incorporate such human intuition in the software for AV, and the absence of such intuition could put pedestrians at risk. Several companies are working on creating software that can “read the minds” of pedestrians–or at least better predict their behavior.

 

Even more alarmingly, there’s new evidence that self-driving cars have difficulty detecting pedestrians with darker skin tones. A study by the Georgia Institute of Technology found that object-detection models similar to those used in self-driving cars detected darker-skinned pedestrians with less accuracy than lighter-skinned pedestrians. The difference was as high as five percentage points.

 

While the scientists weren’t able to use the actual object-models used in AVs currently on the road (companies keep such technology a closely-guarded secret), the study does suggest how human bias creeps into technology. The authors of the study noted that the disparity is likely because the object-detection models had been trained primarily with light-skinned pedestrians and didn’t give darker-skinned pedestrian models enough weight in learning.

 

“The main takeaway from our work is that vision systems that share common structures to the ones we tested should be looked at more closely,” Jamie Morgenstern, one of the authors of the study, reported toVox.com.

Over the past decade, driving has become increasingly safer for drivers and passengers in vehicles yet more dangerous for pedestrians. In February 2019, the Governors Highway Safety Association estimated that over 6,200 pedestrians were killed in 2018, representing a 50 percent increase from 2009.

 

As self-driving cars become more of a presence on the road ,we need to put as much emphasis on making the road safer for pedestrians as we do other drivers.In the meantime, pedestrians should stay on their guard.

 

https://arstechnica.com/cars/2018/07/new-software-gives-self-driving-cars-intuitive-understanding-of-pedestrians/

https://www.vox.com/future-perfect/2019/3/5/18251924/self-driving-car-racial-bias-study-autonomous-vehicle-dark-skin

https://www.ghsa.org/resources/news-releases/pedestrians19

Want To Discuss Your Situation With Our Attorneys? Get A Free Case Evaluation Now

Submit your review
1
2
3
4
5
Submit
     
Cancel

Create your own review

Hyland + Padilla, PLLC
Average rating:  
 0 reviews

No Comments

Sorry, the comment form is closed at this time.