Hit enter to search or ESC to close

Self-Driving Cars Need Geofences Like Flippers Need Water: The Case for Virtual Boundaries

I find it very funny when people say, if an autonomous vehicle has a geofence, it’s a failure. Everything that works in a lot of places had to work somewhere first. Skipping the somewhere part seems like a recipe for bad things to happen.

That’s what a geofence is: a virtual boundary for a real-world geographic area. People have geofences, and so does everything we use. We may not know where they are, we may not even want to admit they exist, but they’re there, and the difference between a good day and a bad day is knowing where they are in advance.

My Roomba has a geofence. So does my 2-year-old daughter, Coco. There are places I trust them to go safely on their own, and places I don’t. Neither are failures. It would be nice if my Roomba could clean my driveway, but Roomba-maker iRobot doesn’t sell street sweepers. They’re too busy making money selling little autonomous robots that clean houses. Likewise, Coco doesn’t know what her geofence is, but I sure do, and I expand it for her a little more every day.

It’s hard enough to build an autonomous vehicle, but it’s very hard to build a safe one. How hard? Trust me, you don’t want to just hand autonomy out willy nilly. A geofence isn’t a limitation; it’s a sign of wisdom.

In Coco’s case, my wisdom.

Before I leave the front door open for her, she’s got to learn some basic rules. I need to see her get around the house safely. Not just once, but many, many times. We have a relationship built on trust. She trusts that I won’t let her get hurt. I trust that she will learn from the occasional bump or gentle fall. As she learns, our mutual trust moves her fence. How far? As far as wisdom allows. Eventually she will have no fence at all — at least not imposed by her parents — and her autonomy will take her anywhere she wants to go, anytime.

Well, maybe not everywhere, anytime.

Because there’s another dimension to freedom, for both Coco and an autonomous vehicle. Engineers call it an ODD, which is shorthand for an Operational Design Domain.

An ODD is the domain in which things are designed to work. It might be a place, a time, a temperature or speed. Maybe it’s two out of four, or even all at once. Everything we use — and especially the things that move us around — has an ODD.

Take flippers. They go on our feet and improve our speed through water. Flippers are GREAT in water. But not all water. Flippers aren’t so great in bathtubs. A pond is ok. Pools and lakes are better. In the ocean, they really shine. You can wear flippers anywhere you want, but try walking on sand more than a few feet. You won’t enjoy it. Try riding a bike. You would look incredibly foolish. Also, it will be very, very difficult.

So what is the ODD of flippers? Water, of course. But only water deep enough to use flippers effectively. Flippers have a pretty simple ODD. Day or night? Doesn’t matter. Temperature? Doesn’t matter.

Planes work the same way. All planes fly. But it would not be true to say the ODD of all planes is “the sky.” Some planes are designed to fly higher than others. Some have de-icing technology allowing them to fly in cold weather. Different planes have different ODDs, and there is no such thing as a plane that can do everything.

Things that are specialized are much, much, much better at doing special things. You want to wear sandals in the snow? Or snow boots in a pool? There’s a reason houseboat racing hasn’t caught on. It’s why houseboats don’t cross the Atlantic. It’s also why a rocket that can reach the moon isn’t ready to go to Mars.

It’s easy to say something can go anywhere, anytime. But it’s hard to go anywhere, safely, every time. Which brings us to the remark that got me laughing about this in the first place: If an autonomous vehicle has a geofence, it’s a failure.

FedEx started deliveries with 25 U.S. routes in 1973. Today it’s The World, On Time. Was it a failure because carriers didn’t deliver internationally on Day 1? 

McDonalds opened in 1940, offering affordable food at high volumes out of a single location. Today it’s in 100+ countries. Was the first McDonald’s a failure? Of course not.

Successful businesses are designed to scale. In the autonomous vehicle sector, that starts with knowing exactly what your vehicle’s ODD is and building a business to fit. It might be a neighborhood. It might be a city. Day or night. Rain or shine. Whatever it is, you’ve got to be ready. Why? Because if your product isn’t safe and predictable, there is no business. Because the reality is that you’ve got to start somewhere.

I suppose one could wait until autonomous vehicles work everywhere before opening doors for business, but that would make no sense. Competitors are out there, so why wait? You go to market with a great product, and if designed right, scale should flow from there. As the business evolves, so will the ODD.

There is no magic to growth except execution.

Meanwhile, Coco’s ODD is growing, too, and her safe geofence right behind it. Another room last week. The edge of the patio this week. Next month, who knows? I’m impatient to see her go anywhere she wants, but right now I’m happy with her getting just one place safely, every time. 

Which is exactly what I want from autonomous vehicles, too.

Choose your lane

How Autonomous Vehicles Distinguish Between Bicycles and People Who Ride Them

How Autonomous Vehicles Distinguish Between Bikes and People

When it comes to how autonomous vehicles see the world, humans come first, literally. Autonomous vehicles (AVs), like the kind operated by Pittsburgh-based Argo AI, use Machine Learning to detect and classify the objects in their surroundings, identifying people...
Why The League of American Bicyclists is optimistic about autonomous vehicles

Why a Leading Cycling Advocacy Group Is Optimistic About Autonomous Vehicles

As autonomous vehicle use grows, AV companies and the League of American Bicyclists are collaborating on how to ensure cyclists and motorists can share the roads safely, even if the “motorist” is artificial intelligence software. As part of the...

Self-Driving Is Arriving Right On Time. Just Like Ice Cream Did

Seven years ago, I was a self-driving skeptic. Not of the technology. Of all the “experts” promising autonomous vehicles would be everywhere by 2020. You didn’t need to be Nostradamus to know that was ridiculous. All you needed was...
Illustration of a futuristic parking deck turned into a mixed-use space, with AVs driving by

How Autonomous Vehicles Could Help Transform Parking Lots

Researchers say it’s likely that autonomous vehicles (AVs) can help reduce the need for parking lots, opening more room for grass and trees and other elements of nature. It may not seem like it when you’re circling the block...
An illustration of an Argo autonomous vehicle in teal against a backdrop of buildings, a bicyclist, and research papers

7 Big Breakthroughs From Argo Research at CVPR 2022

The 2022 Conference on Computer Vision and Pattern Recognition (CVPR 2022) is nearly here. Thousands of computer scientists, software engineers, and researchers from around the globe will gather in New Orleans to review and discuss their latest work in...

Researchers Predict the Future With Lidar Data

Researchers at Carnegie Mellon University’s Argo AI Center for Autonomous Vehicle Research, a private-public partnership funded by Argo for advancing the autonomous-vehicle (AV) field, say they have come up with a way to use lidar data to visualize not...

Must Reads