If driverless cars were to emerge only after all infrastructure has been uniformly upgraded and every possible hazard anticipated, it might be a long wait. An override to these problems is autonomous vehicles being connected to a network–and each other–and constantly be “educated.” In Steve Ranger’s ZDNet interview with Jim McBride of Ford’s driverless division, the latter addresses this issue, promising that shift from driver to driverless will “not terribly dissimilar from [the shift from] horses and carriages going to cars.” An excerpt:

Question:

What are the big technical challenges you are facing?

Jim McBride:

When you do a program like this, which is specifically aimed at what people like to call ‘level four’ or fully autonomous, there are a large number of scenarios that you have to be able to test for. Part of the challenge is to understand what we don’t know. Think through your entire lifetime of driving experiences and I’m sure there are a few bizarre things that have happened. They don’t happen very frequently but they do.

Question:

How do you build that kind of intelligence in?

Jim McBride:

It’s a difficult question because you can’t sit down and write a list of everything you might imagine, because you are going to forget something. You need to make the vehicle generically robust to all sorts of scenarios, but the scenarios that you do anticipate happening a lot, for example people violating red lights at traffic intersections, we can, under controlled conditions, test those very repeatedly. We have a facility near us called Mcity, and it’s basically a mock-urban environment where we control the infrastructure. While you and I may only see someone run a red light a few times a year, we can go out there and do it dozens of times just in the morning.

So for that category of things we can do the testing in a controlled environment, pre-planned. We can also do simulation work on data and, aside from that, it’s basically getting out on the roads and aggregating a lot of experiences.•