I assume many of these seemingly odd efforts by big companies are to help fill their patent portfolio.

That is also a very reasonable hypothesis. It would be interesting to know the economics in this case, if that turned out to be true. Spinning up an entire automotive engineering R&D division seems like a very expensive way to generate patents.

> Spinning up an entire automotive engineering R&D division seems like a very expensive way to generate patents.

What's the alternative? Real patents usually come from real work, rather than thought experiments.