Today, Episode 81! The University of Michigan’s Transportation Research Institute (UMTRI) has an idea to help lower AV testing costs by utilizing remote human operators; Amazon finally invests in an autonomous vehicle startup; given 5G’s limited range, how will it be deployed to benefit AVs without coverage dead zones? (Spoiler alert: I don’t know!)
Air traffic control for AVs?
So this is interesting. According to this article, the University of Michigan’s 50 year-old Transportation Research Institute (UMTRI) is considering relocating human backup drivers to remote operating facilities. The idea is to help keep costs down, because instead of having n backup drivers for n cars, you would need, say, x backup drivers, where x < n. The idea is that the cars would be able to anticipate 10-30 seconds down the, er, road, if they might need human intervention, and then notify a remote driver to help take over.
Amazon invests in AV startup
Spoiler alert: Amazon wants in on the AV game. To that end, they just joined a $530M round into Aurora, an autonomous vehicle group not building their own AV, but rather working with others like VW, Hyundai, and China-based Byton.
How will 5G be deployed?
Depending on whom you ask, 5G is either essential for a connected, autonomous future; useful but not necessary; or an utter waste of resources. The promise is the low, 1ms latency of 5G will enable lag-free remote vehicle operation; cloud-based data processing, and more. This contrasts with the relatively eternal 30-100ms or more of existing 4G LTE systems (and even wired networks). The problem, however, is that 5G, due to its far higher frequency (shorter wavelength), is extremely susceptible to interference, and thus has a far shorter range than, say 4G LTE. So you’d need a tremendous number of 5G radio towers to ensure no dead zones. But we can’t even accomplish this with 4G. So how is this going to work?
Like this podcast? Please support it! Leave a 5-star rating and review on iTunes!