fbpx

loading the future of Autonomous Mobile Robots

+

Reach out and we will get back to you!

    28.06.2024
    arculee M Electronics: The Story of Efficiency and Fast Development

    The electronics of Autonomous Mobile Robots (AMRs) play a key role in determining their functionality and efficiency. The same goes for arculees, our AMRs. Join Tobias Schwering, Andres Magallanes, and Jolly Jacob from the arculus electronics team as they explain how the arculee M, our latest product, builds on the electronics strengths of its predecessor, the arculee S, but also offers several advanced features.

    The Evolution of Electronics: From arculee S to arculee M

    Can you describe the evolution of the electronics aspects from the arculee S to the arculee M? What were the key improvements or changes made?

    Tobias: One of the main goals for the arculee M was to ensure fast development. To achieve this, we reused as many components from the arculee S as possible, thereby accelerating the process. We identified key parts that could be reused, starting with the RCU (Robot Control Unit).

    The RCU, initially designed for the arculee S, was adapted for the arculee M with some additional functionalities to support new use cases. For instance, we incorporated a safety concept for ramps. We have not designed the S to navigate ramps, but the arculee M can go up to 10-degrees on the ramps. So, this was a big thing. Besides adding these few interfaces, we made general improvements to the RCU. However, for Andres, the harness presented different challenges.

    The arculus  RCU®, designed for arculee S was also adapted for the arculee M  

    Andres: The harness plays an important role in the mechanical and electrical design of a robot. There was a major change here from the arculee S to the M. In the S, we used a single long harness, which, if you put it on the table and extend it, will probably be three meters long. In contrast, the arculee M features a more modular approach. The modular approach means splitting the harness into different modules. Therefore, we easily replace, handle, and manufacture them, simplifying the assembly process during production.

    However, the harness differs not only in modularity but also because of the new components it supports. We have introduced different devices, such as upgraded scanners and new data communication interfaces, marking a substantial evolution from the arculee S.

    Tobias: Perhaps we should clarify that when we mention “scanner,” we refer to the safety laser scanner. This type of scanner ensures our robot can operate safely around humans, making it functionally safe. Additionally, we use the safety laser scanner for navigation purposes.

    Jolly: Another aspect to consider is the design. We had already tested many RCU modules with the arculee S and found them efficient and high-performing. Therefore, we retained these modules in the arculee M without compromising efficiency or performance. We took the proven modules from the arculee S, integrated them into the new RCU, and made necessary modifications to suit the new robot.

    Spools of harness wire
    The harness wire spools at the electronics labs

    The Technical Aspects of the arculee M Electronics

    What are some of the most significant technical innovations in the electronics of the arculee M?

    Jolly: The camera used in the arculee M would be an innovative technology, and the brake control board, with its sensing method, would be a technological advancement, too.

    Tobias: We updated our safety scanner system to provide more precise laser data. This enhancement improves the entire software stack, leading to better localisation, system navigation, and more robust performance.

    Jolly: Another important change, though not a major technical one, is the modification of the LED light. Previously, it was pretty large and used a lot of space. We have now significantly reduced its size to less than half, making it more efficient and directing the light precisely where needed.

    Andres: In this new project, we took a different approach to designing the harness using computer-aided design (CAD). Previously, we would create a prototype harness for the robot, document the length, estimate wire coupling lengths, and then compile a bill of materials based on what we used or had available.

    While we had electrical documentation specifying the connection procesdures, we lacked detailed documentation on how to build the harness. Now, with the CAD approach, we have integrated the harness design into the 3D design of the entire Autonomous Guided Vehicle (AGV). This method is easier for manufacturers to maintain and handle, allowing for the direct creation of production drawings.

    Jolly: The new approach allows us to easily remove parts for testing or modification with the arculee M. We had not implemented this method in the arculee S. It’s a valuable addition that we can use in all scenarios, including testing during production and maintenance.

    Tobias: It was mainly designed for production so that we could pre-assemble modules. That was the main idea, I think. But then we realised that it also has several other advantages. So now you can just exchange parts. For example, if something breaks in the field, we can just ship one part there and exchange it.

    Andres: Now, we are more interested in making it easier to produce, assemble, and maintain and repair on the field. So, if one reset button is not working, we can just remove the panel, replace it with a new one, and then investigate what the problem is with this button.

    Tobias: Well, a robot is just such a complex thing. So we call it always the integration phase – the first time a robot is built, and then we put everything together. And there were tests upfront. But your tests are never complete until you have a complete robot. The completion of this, with every single robot that I saw, usually would take a long time. However, with our modular approach with the arculee M, more people could work on it in parallel, which sped up this process.

    After all, that’s one of the key points where speed and efficiency in development are required.

    Two engineers working in the arculus lab
    Andres and Jolly are busy working in the arculus electronics lab

    How have power management and efficiency been improved in the new electronics setup?

    Tobias: The biggest addition was the deep sleep feature, which allows the robot to reduce its power consumption to extremely low on a remote command. Also, the robot can restart itself — it all comes from the fleet manager!

    Except for that, I think the power consumption and efficiency were already pretty great in arculee S, and arculee M is now also relatively efficient. We have our completely self-designed inverter system, which is very power efficient.

    Jolly: One improvement is to speed up the lift timing. So we just increase the speed of the lift, and in that way, we save some operation timing. Also, we just changed some circuitry with proper DC converters, ensuring more efficiency. Initially, we had designed for more power, but we don’t require it with arculee M. So, we redesigned those sections to save some power.

    Tobias: The arculee M weighs nearly twice as much as the arculee S. It can lift more, and move around more with less power consumption.

    The Challenges of the arculee M Electronics

    What were some of the biggest challenges faced in developing the new robot, and how did you overcome them?

    Tobias: One of the biggest challenges was the time frame. We wanted to develop arculee M very quickly. However, we had to make changes to the RCU, including adding new interfaces. For example, we introduced a new front camera and 3D camera system, which required an additional Ethernet interface on the RCU.

    We had to push out a new revision of the main board fast. Since it’s a large PCB (Printed Circuit Board), we had to order, assemble, test everything on schedule before starting real integration. So, fitting this into the development cycle while ensuring enough time for proper testing was a significant challenge.

    Andres: Also, this braking board is a very interesting system because, according to the current, you can detect the opening and closing of the brake. The industry already uses this principle, but not so widely. And yeah, implementing it here was a new experience overall.

    Jolly: We quickly assembled our new PCB for the brakes and other components, tested it, and debugged everything to ensure it worked as expected. The tight timeframe was, thus, one of the challenges we faced, requiring us to keep working at full speed.

    Tobias: But overall, we did three revisions, and had to add only one connection. Over these two revisions, we added two resistors. Besides, we have now built a technology base that we can use in the future, for example, for electronics module development.

    A quote from Tobias Schwering, one of the engineers at arculus
    Tobias explains the importance of functional safety when making design and electronics decisions for an AMR

    Integration of Mechanics with arculee M Electronics

    How do the robot’s mechanics and electronics integrate to enhance performance and reliability in the new model?

    Andres: Well, I think the harness makes a great example. Moving from a big, fully integrated harness to a modular approach required us to work with the mechanics department. Of course, the design initially comes from the electronics, which places all the components you need there, estimating wires, cross-sections, cables, and all the signals that need to go. But as I mentioned before, it’s now fully integrated into our robot’s CAD system, which is where the mechanics enter the picture.

    Tobias: The modular approach made achieving electromagnetic compatibility (EMC) more challenging. We must adhere to some limits regarding how much a robot can radiate. Of course, we’ve managed to do it with the arculee S – we have passed the official certification in an accredited lab. But with arculee M, everything was larger, and there were more connectors where the wires were pulled apart for easy access and more interconnections in the single parts. So, it’s not one single structure that works as a shield anymore. This meant collaborating with the mechanics department to develop a system that fulfilled the modular approach’s requirements without ruining the robot’s EMC capabilities.

    Jolly: There were some other areas where electromechanical integration was necessary. For example, if we had to place a PCB in a different part of the robot for lighting, we needed to fix a position for that. The mechanics team would help place it at a proper angle and in a certain way. Similarly, mounting the battery was also where the electromechanical integration was crucial.

    The arculee M Electronics: Lesson Learned

    Let’s talk about the lessons learned. What lessons learned from the development and the deployment of the arculee S were applied to the new model’s electronics?

    Andres: One thing we have learned while developing arculee S was that if you want to be fast with the design, you need to parallelise the work with other teams. You cannot block the next team’s work; it will never work, and we implemented parallel working for the arculee M.

    Tobias: But it goes both ways because we also learnt what strategies and components used in the S were really good ideas. For example, the drive system and our self-developed inverters turned out to be more or less one-to-one usable in the arculee M. So, of course, there were lessons about things that were not so well with the S, but also about what worked and can be used as it is in the next robot.

    Andres: We just changed some parameters because we used different motors with different physical characteristics, but that was all!

    Tobias: Yeah, but only for lifting. The drive motors we used for the arculee M only have a stronger brake. The rest is the same. So, that was only for the ramp. As we discussed, the arculee M is heavier, can carry more, and can stay on ramps. So, it has to be stable and shouldn’t start moving them. This is why we had to beef up these electromechanical brakes a bit. However, that was not in the scope of the arculee S.

    Moreover, we added the brake controller board, a supervision feature. The idea was to ensure the brakes are actually were stronger and working before we introduced them to the driving into ramps! Otherwise, that might be dangerous. Since functional safety is a major topic in many design decisions, including electronics and AMRs work in close collaboration with humans, we must ensure that nobody gets hurt working with a robot.

    Robot testing area at the arculus new office
    The spacious testing area at the new arculus’ office

    Jolly: Also, we tested this with an actual load, not in a simulation. So we put the real load on and applied the brake while moving. Then, we performed all the tests immediately to ensure everything was working perfectly.

    Tobias: Moreover, our basement room at the old office wasn’t large enough for the brake test. You know, to bring the robot with a weight of 1.75 tons – actually, we even tested with overload, so 2.05 tons, to full speed and then brake. But we managed to find suitable test halls then, thanks to Jungheinrich, our other company.

    However, with the new office, one of our biggest concerns was how much the floor could take. Thankfully, we have access to a great testing area now!

    This is all about the arculee M electronics. Watch the entire interview here!