Why Weather Prediction Is Critical For Self-Driving Vehicles
“I’m an optimist,” said Harold Wilson, ”but an optimist who carries a raincoat.”
There aren’t a lot of certainties in life, but one thing’s for sure: a real expert will always tell you what they don’t know. I just wrapped a podcast with The Weather Channel’s Jim Cantore — one of the most popular meteorologists on the planet — and it was a great reminder that no matter how much time, money and technology one throws at a problem, a little common sense is still necessary.
Can we predict the weather with 100% accuracy? Of course not. But do we need to? If there is a 90% chance of rain, I bring an umbrella. I might even bring a raincoat. Or I might wait until the rain stops. But no matter what the forecast is today, common sense starts with having bought a raincoat and umbrella yesterday.
Which brings us to a huge misunderstanding about self-driving vehicles, specifically robotaxi and commercial delivery services.
A lot of people seem to think self-driving is a software problem. Finish the software and cars will just magically drive themselves. But that doesn’t make any sense. What if it rains? Two things have to happen for a self-driving vehicle to operate in the rain. One, it has to be properly maintained. Wiper blades intact. Wiper motors functional. Headlights good. Tires properly inflated. Secondly, it has to know when not to go.
Just like people, self-driving cars won’t be able to operate in all conditions, like a hurricane or a tornado. But unlike many people, a properly designed self-driving system will be aware of those conditions and make smart decisions to avoid them.
“There’s no constant to human behavior,” Cantore said, “because everyone behaves differently. They assess their own risk.”
A willy-nilly approach to assessing risk ain’t going to cut it if you want to operate robotaxi or commercial delivery services. Customers and businesses alike need to know a robotaxi will deliver goods, groceries or medicine on time.
And if the weather looks bad, not in one hour but in four or six, arrangements must be made. A delivery may need to be rescheduled. A human driver may need to handle that route.
Either way, what people call reliability — and businesses call uptime — is the same thing: trust. The better your forecasting, the better your customer service.
Technology isn’t magic, it’s a tool. And done right, it gets better every day.
Hear more from the No Parking podcast here.