4 ms·
One of the main problems is that we have many different software teams trying to solve the same safety feature driven functions in many different ways with many
by cyrks 5y ago
One of the main problems is that we have many different software teams trying to solve the same safety feature driven functions in many different ways with many different solutions, outcomes, and decisions along the way. This leads to a very fragmented base of software with varying levels of safety, none of which is tested to the same standards. The result is that consumers don't know of the car they are buying is truly safe at all
- judge2020 5y agoThis is something agencies like NHTSA and/or private companies like IIHS are supposed to solve, but most active safety technology testing is just ‘does it work’ in very easy scenarios - the test suite needs to include many environments and tough conditions if it wants to evaluate these systems more broadly (The only issue being when car companies optimize for those specific tests, or ‘when the measure becomes a target’).
- Red_Leaves_Flyy 5y agoAnyone want to fund me? I’ll test the every living heck out every smart car in existence in every condition from perfectly flat straight drag steps to dirt roads in the mountains with a foot of snow and white out conditions, to unmarked pot holes riddled roads in torrential down pours. I’ll also test the usual stuff like does the car detect the kids playing in traffic, or the dog darting after the squirrel. I’ll video the test from numerous angles, cut in awesome music and sounds effects and upload everything to the internet. I’ll also write long articles about the intricate nuances of each vehicles responses to the tests and create scoreboards for categories like least likely to run over children, or most likely careen off a bridge if there’s a sign at the perfect angle.