5 ms·
Most self-driving is about avoiding collissions, and signalling intent, especially when streets are narrow and there's merging or shared use. The physics of car
by teyc 4y ago
Most self-driving is about avoiding collissions, and signalling intent, especially when streets are narrow and there's merging or shared use. The physics of cars, people, bikes and kids around roads are well understood (acceleration, velocity). This can be simulated, and a game engine can generate data for virtual sensors to be trained. There's no reason to require time on the road.
- cptskippy 4y agoBut you'll never be able to come up with all of the possible scenarios to simulate. What Tesla has demonstrated is creating virtual scenarios where they can dynamic adjust all factors (light, weather, traffic, etc) and base them off real world situations they've encountered where their Model failed.
- bastawhiz 4y agoMaybe not manually, but surely you could develop an adversarial ML model that quickly and concurrently tests scenarios.
- cptskippy 4y agoWhat data is that model based off? Tesla has the data based on real world failures to build that model. Does anyone else?
- bastawhiz 4y agoYou can't discount all the data they Waymo has collected over nearly a decade or the scenarios they've manually created. They also have the world's most complete map and spatial dataset, which could easily be extended to create a model that creates tricky roadways. Stimulating obstructions or hardware failures doesn't require very much data at all. If you are modeling scenarios like a game engine, a "discriminator" model isn't necessary: you just check whether a simulation doesn't result in a crash.
- cptskippy 4y agoI'm not discounting their data, I just think Tesla has so much more. If you were looking at just those opted into the FSD Beta you have a larger fleet actively running the model with feedback loops capturing every failure. But cars without FSD are still running the model and capturing data as well.