Super-MP Startup: Avoid 6 Critical Setup Errors | Marvelmind

0:00 Hello colleagues. Let's continue helping with examples of typical customer deployments. So this is the setup room: Super-Beacon, Super-Beacon, Super-Beacon, and another Super-Beacon. The problems that customers face are the following: first, the table of distances doesn't build automatically—it stays red. And then when they tried to make a ROS integration, it didn't work. So let me first mention ROS. Of course, as we repeat many times, do not move further if you didn't succeed with the previous step. So we have a page about that, we have a video about that. So it means that you need to achieve perfect tracking first, and only then try to integrate. There is no point to start

1:00 integrating when you don't have perfect tracking. So that's the second step. Let's discuss the first step: why would the table of distance not build? What could be the reasons? Again, we don't have full details, but according to the email that we received, the software has been updated. So the first check is, of course, the first check: make sure that the software was really updated with the right software. Non-Inverse Architecture—then all the beacons must be in Non-Inverse Architecture, and the model must be in Non-Inverse Architecture. If it's Inverse Architecture, then all the beacons and the model must be in Inverse Architecture. Now, presumably everything was done right—it looks like it—but that's what must be checked. Second, and it's not clear whether it was done or not, is pressing default buttons. We keep repeating that default buttons must be pressed

1:59 for all beacons, each and every one, when the beacons are software-updated, and for the Modem. Because if you don't press the default settings, then some settings would be used. What are those settings? Well, we are producing in the factory using some factory settings. The software which was used had, you know, probably been from months ago, and the software between releases may change drastically. So it meant that before you do any testing, before you do anything, you must come to a single configuration which is the same for all beacons, and at that moment that is achieved by updating the software from the same software pack and pressing the default button. So it wasn't clear whether the default button was pressed or not, and it could be

2:58 one of their problems. Let's discuss the next one, which is potentially linked indirectly with the default settings. So these are the Super-Beacons. Super-Beacons emit 360 degrees horizontally and 180 degrees vertically—so basically a hemisphere. It's for transmitting. There are five transducers transmitting in all directions horizontally and one vertically. Altogether they cover this hemisphere. For reception, they're using the microphone. The microphone has also 360 degrees and 180 degrees reception diagram, but their signal strength or the sensitivity from all directions is not the same.

3:57 So obviously, the sensitivity from the front direction is much stronger than the direction next to the horizon. And in this case, it's not only next to the horizon—it's really to the horizon. That could be not a problem if you don't have such a close distance to the ceiling. What may happen—again, we don't know for sure because more details are required—but what could have happened: the very strong signal is emitted and then reflected. The strong signal is emitted, but their reception from this direction is significantly weaker. Reception, no transmission, but reception by this beacon would be significantly weaker than from this, because the sensitivity diagram of the microphone. And what happens? Yes, the system is trying to get the front, or the first signal, which is the

4:55 direct signal. It's always first. But in presence of such a strong reflected signal, the system could have confused and decided the front signal is actually some sort of noise. It could be even worse if the default settings were not pressed. In this case, the situation is almost unsolvable. Why? No, for example, in our factory, only their front sensor—the front transducer—is enabled. So the only front transducer is transmitting, and the side transducers are simply disabled. We don't know this, but they could be. In this case, the signal to this wouldn't even go strong—only this front signal would be strong. So only reflected signal would be strong. So there was no even chance to get it. Maybe that was the case, maybe not. We don't know, but the solution is very, very simple: always

5:54 press default on each and every beacon and on the Modem. And to be even more sure, after you press the default, check that the sensors are actually in green, so they are enabled. What could be done if that is not the case? Well, we suggest not to place them like this, so they would look to the horizon, but to place the beacons a bit more vertically so they would look to the center where you want to be tracked—with the front sensor and the microphone. So the microphone would be looking to the center. Now the microphone is looking at their ceiling. If you need to cover all this room, then our recommendation would be different. Our recommendation would say, okay, put this on the wall like we recommend in the operating manual, like we recommend

6:54 in the placement manual. So this is possible, doable. But in this case, also we recommend putting this on the ceiling, not like this. But in order to build their table of distance, the beacons must hear themselves very, very well. If you could put this on the ceiling, then the distance to the reflected surfaces would be significantly larger. First of all, and second, it would be such a flat—so it means that there will be—it wouldn't be such a perfect reflection. There would be some sort of dispersed reflected signal. So it means that the problem—if that is the problem due to reflection from the ceiling—it wouldn't happen if their beacons would be placed on the wall. It certainly wouldn't happen if the beacons would be placed on their walls or on the ceiling. It wouldn't happen if the beacons would be placed in the wall like we recommended. So it meant that there's always kind of a set of small mistakes. But when there are

7:53 several of them, then, you know, they accumulate and accelerate each other, multiply each other, and then there's a problem. So, solution: press default settings. So make sure that they are pressed. Really, second, place the beacons as recommended in the operating manual on the wall or on the ceiling, not like this. I remember that there could be some reflection, and if your direct signal is weak, then the reflective signal may simply overcome. And then the last one is also very, very simple: if you cannot succeed with a more complex map—like, you know, four stationary beacons—always simplify. We have a page about that, we have a video about that on typical mistakes. Do 2D tracking with two stationary beacons. Place them on the wall. One mobile beacon is in your hand, and walk around. Perfect tracking. Great! Then increase the complexity. Do not jump

8:51 immediately to something more complex than you have achieved already. Hopefully it's helpful, hopefully it's understood, and if any additional questions of similar sort, send us. We'll be able to produce videos on examples, and it would help in this and similar cases. Thank you very much.