Niels Jul Jacobsen is the chief strategy officer of Mobile Industrial Robots (MiR), a leading manufacturer of autonomous mobile robots based in Odense, Denmark. He has worked in robotics since 1989 and has been a key figure in Denmark’s booming robot industry. When will AI, applied to data gathered by vision and sensing technologies, begin having an impact on autonomous mobile robots? It’s just emerging now. We’ve seen it in certain products over the last few years but not yet in real-world applications. It’s still coming from research universities and small startups, but it hasn’t been put into real production yet because we don’t think it’s stable and robust enough. But it’s something that we’re working hard on. The ability for the robot to have a better perception of what’s going on is very important. In the near future, robots will be able to see what kind of objects are in front of them using standard technologies from deep learning and image recognition that are already available today. And based on this, the robot will be able to make better choices on what actions it should take given the obstacle, person, or truck in front of it. Today we have to restrict robots a bit more in order to make them safe enough to operate in manufacturing environments. With advances in vision systems coming, we foresee that robots can move faster in environments where there are a lot of people or other moving obstacles, because the robots will be able to see what’s going on, and, like humans, find a way to navigate around them. To what extent have technological advancements in collaborative autonomous robots mirrored those in the automotive industry? The push to have autonomous cars is moving vision and AI forward very quickly, and we have been able to utilize the technology developments that have gone on in the field. Despite this progress, the technology is still not quite ready yet. We still need to see whether autonomous vehicles can fulfill their promise. There are many situations, such as on the highways, where it’s obvious driving can be automatized, but moving outside the highways to the city environment is really a big challenge. It’s easier with our mobile robots because they don’t move as fast and it’s easier for them to react. What’s next for AI, vision, and collaborative mobile robots? Soon you’ll see mobile robots being introduced on a larger scale, in all types of factories. At least from our point of view, it’s something that will occur in the next two years. But it will be done in a gradual way so people can get used to the robots and developers can learn from them. AI will be put in as a system to enhance the robots’ abilities, but if AI were to fail for some reason, we could still fall back on our current solutions. If AI works as intended, you get more of it, but if it doesn’t, it’s not a dangerous system.
