David Vedvick

Notes

July 7th, 2022

Simple things Tesla could do with their overpowered CPUs and GPUs to improve the driving experience:

  1. Range estimation based on weather, simply requires data (of which they have plenty) and regression models, using temperature, windspeed, and altitude changes as an input. To make it fancy, they could adjust it based on predicted temperature along a route, but while technically feasible, this might be a major pain. To make it even fancier, they could add component lifetime as an input. All of this is hard with traditional rule-based systems, but should be natural for machine learning. Supposedly, Tesla has a supercomputer creating advanced ML models that could easily create this kind of model.
  2. Simulate summer weather braking during winter weather
  3. Predicting component lifetime, detecting car problems ahead of time

Note posted on Wednesday, July 6, 2022 7:17 PM CDT - link