Among the bevy of problems self-driving cars face, bad weather is the most significant of them, not that human drivers can handle it too well. To mitigate this, Waymo hopes to convert their autonomous robotaxis into mobile weather meters.
The Alphabet company has announced that the latest version of the sensor array on its autonomous vehicles – using a combination of cameras, radar and lidar – is capable of measuring the weather conditions the car may encounter, specifically the intensity of raindrops (or lack thereof), as well as fog. It would turn the vehicles into, as the company puts it, “mobile weather stations.”
This doesn’t mean you’ll see a Waymo car reporting the weather on your local TV station any time soon, but it will help the robotaxis make real-time decisions to adapt to weather conditions on the ground. To begin with, it is being tested in Phoenix and San Francisco, two very different climates.
This approach could potentially offset the limitations of relying solely on weather data from airport weather stations, satellite and radar sources and provide more local, pragmatic data, in this case what’s happening directly in front of the car. That can come in handy if a black cloud follows one of them like a cartoon.
But since the sensors ostensibly turn the vehicles into amateur meteorologists, Waymo can also use the data to create real-time weather maps on conditions such as the advancing coastal fog and light drizzle that a radar might miss.
The technology itself is clearly as in its infancy as self-driving cars, and we’re nowhere near the kind of uninhibited autonomous technology that movies love Minority Report and i robots led us to believe we were around the corner.
But when you look at the number of road accidents involving human drivers who mistakenly think they know how to drive in rain or snow, it seems that extra help in this area can do no harm.