Is the industry robot-ready? interview with Dr. Christian Wurll

Dr. Christian Wurll is a professor of electrical engineering and automation at the University of Applied Sciences in Karlsruhe, Germany. Dr. Wurll has extensive experience in the study and application of robotics and automation. Before entering the academic world, he held executive positions at companies including Swisslog Automation, Grenzebach, and Kuka Robotics.

Dr. Wurll studied control theory and robotics at the University of Karlsruhe, Germany, where he earned a master of science degree in electrical engineering and a Ph.D. in computer science. He has published extensively on robotics, vision, and search algorithms. Dr. Wurll spoke with DC Velocity Senior News Editor Ben Ames at the MHI Fall Conference in La Quinta, California.

Q: Supply chain managers are facing many challenges. They are dealing with labor shortages and demands for ever-faster fulfillment. The robots are coming, which may help to ease the labor crunch. But are they really ready for 24/7 use yet?

A: I have implemented many systems and applications in logistics warehouses, but still using the traditional industrial robot arms, which are basically caged in, in order to protect the people running the applications. However, there is a new trend coming with all the cheaper robots—we call them "collaborative robots." They are designed and developed to run without a safety fence, but that requires that these robots be operated in accordance with certain rules and safety precautions.

What we observe in the European Union is that the machine laws are pretty tough. [Regulators] have set the forces applied to, say, a collision with a human pretty low, which means that you have to run these applications at very, very low speeds. But that can make it difficult to meet productivity goals. You want to achieve high throughput rates, but you can't achieve that running at a very slow pace. That means that it is a hindrance to actually deploy these robots in the numbers the industry is really looking to do.

Q: For a return on investment, you really need to ensure that the robots compare favorably to a human who's doing the same work. Robots might be more reliable—they're not going to take sick days, for example. But if they are running slowly, then the throughput just isn't there?

A: Exactly. It is kind of crazy how many new suppliers for robots are showing up at these industrial fairs, both here in the United States and in Germany. But if you really look carefully into these various vendors, you can still see some differences. Not every robot type is really suited yet to run 24/7, in my mind. They look nice. They are actually well-designed, but if you compare them, you will see there are huge differences. So really, you have to be careful about what type of robot you are selecting, and you certainly have to run some tests.

Q: So, it sounds like there are a number of challenges. The safety restrictions, the reliability, the speed, and also the reality that some of the collaborative robots, or cobots, are not actually operating in a fashion that's so collaborative yet. Is that right?

A: Yes, that's right. Then, keep in mind that you have to look at more than just the robot itself; you also have to look at the gripping technology. That also has to be collaborative. Even if your robot and your gripper are designed to avoid harming the operator, you still have to watch what kind of work piece you are actually handling. If the work piece has sharp edges, then you still have to come up with a solution for avoiding a collision between the robot and the operator.

Q: So, the gripping technology is one aspect of the warehouse-robotics development picture. But I believe that in some of your writings, you've suggested that pallet moving is going to be one of the up-and-coming applications for robots?

A: Absolutely. When we install systems with large integrators, I often see many pallet conveyor styles being implemented. Sometimes, there are miles of conveyors. I think there is a trend coming up with replacing these pallet conveyors with mobile robots. We can see it with the systems now [used] at Amazon Robotics. They have been a game-changer in the industry. Now, robotic technology can also be deployed to heavier items, like pallets. I think they are definitely much more flexible [than conveyors], and you don't have to bolt something onto a floor, which allows you to be more creative in your layouts.

Q: As these applications develop, particularly with regard to pallet picking, it seems like one of the enabling technologies to accelerate that might be 5G wireless technology, because really, they always have to have a data link in order to work, right?

A: That is true. We are closely watching what 5G will bring to us. Certainly, you have to have the infrastructure first. From what we are looking into, it feels pretty promising. You can connect lots of sensors and actuators to your automation system and start communicating in real time with these devices. That allows you to actually come up with some newer, smarter solutions requiring less cabling and reducing costs in terms of commissioning those systems. That is an interesting trend, and it definitely will change certain things. Look at mobile robots, for example. You always need the navigation principles and methods—so you can maneuver from point A to point B inside a warehouse, for instance. The normal approach is using some sort of "SLAM" technology.

Q: You are referring to simultaneous localization and mapping?

A: Yes, exactly. But it all depends on usually a leader sensor acquiring the data, and with 5G coming up, you actually have an exact position where your object is. So, you may not even need those sensors anymore, because in the communication protocol, you can actually detect where you are. That means that you can more or less simplify your navigation concepts—there's no need for a map anymore. I think that definitely will change the way we integrate mobile robots.

Q: That would be a different approach. And yet the 5G technology is something that's really not in the control of the vendors or platform developers. They have to do that in conjunction with the telephony providers. Is that right?

A: Absolutely. There are some approaches now being discussed in Europe in which, let's say, the OEMs request their own frequencies. They have to be somewhat independent from the telephone suppliers. I think in Europe, we just awarded all these contracts recently, but certain frequencies are specified for these OEMs. Then they can basically be running on their own.