Incremental Map Building for Indoor Positioning | Marvelmind

Installation & Setup

Incremental Map Building for Indoor Positioning | Marvelmind

▶ 16:00
📅 2022-08-20

🔗 Watch on YouTube

For more information, please contact: info@marvelmind.com

Video Overview & Technical Details

This step-by-step guide demonstrates the optimal approach to building complex maps for indoor positioning systems. Start with a simple 2D NIA configuration using one mobile beacon, then progressively increase complexity while maintaining perfect tracking accuracy at each stage. Learn why gradual expansion prevents common deployment failures and accelerates time-to-value for autonomous robots, drones, and forklift tracking applications.

Transcript

This step-by-step guide demonstrates the optimal approach to building complex maps for indoor positioning systems. Start with a simple 2D NIA configuration using one mobile beacon, then progressively increase complexity while maintaining perfect tracking accuracy at each stage. Learn why gradual expansion prevents common deployment failures and accelerates time-to-value for autonomous robots, drones, and forklift tracking applications.

0:01 Hello, colleagues. Let's discuss today how to build complex systems from the basic one, and the key message is very, very simple: go step by step. Don't jump immediately to the final configuration. Don't jump over configuration; increase the complexity one step at a time. So that's the key message of the whole presentation. But let's bring examples. You want to build a complex system to cover a warehouse. You want to track many people. The warehouse will have many submaps, and you want to track a few robots, for example, or forklifts. That's great. Eventually, you want Inverse Architecture, and you want to build this configuration. But don't try to build it right away. Most likely, you'll fail because there

1:01 are so many ways something will go wrong. We recommend to start from the most simple configuration, which is Non-Inverse Architecture. Remember, you want to build Inverse Architecture 3D and multiple submaps. Okay, fine. It will be, but later. Build a basic Non-Inverse Architecture with just two stationary beacons, one mobile beacon. Position stationary beacons—like we really recommend—just two stationary beacons on the wall. Achieve perfect tracking for one simple Super-Beacon, for example. If you have perfect tracking, that's it. That's done. Move to the next stage. Since you want to build Inverse Architecture, your next stage would be Inverse Architecture 2D and one mobile beacon. So effectively, if you deploy the system based on Super-Beacons, you do not change anything

2:00 except for updating the software from Non-Inverse Architecture to Inverse Architecture. And of course, you need to check. And this is why we do not recommend Inverse Architecture right from the beginning. You need to check that your stationary beacons actually have two different frequencies. Because remember, for Inverse Architecture, the stationary beacons are emitting ultrasound. So in order to receive ultrasound at the same time and be able to distinguish between ultrasound from one station because another station begins, they must emit different ultrasound frequencies. So this is why you take the same physical Super-Beacons. They must be already on two different frequencies, for example, 19 kilohertz and 25 kilohertz. You update the software from Non-Inverse Architecture to Inverse Architecture. You update the software to your mobile beacon to Inverse Architecture, and you update the software

2:58 to your modem to Inverse Architecture. As soon as you've done that, and of course, remember to press the default button, you'll have exactly the same physical configuration, but it will be now Inverse Architecture. Check the tracking is fine. If you have it, that's great. So you're already on the next stage. What would be the next stage? Well, you have multiple options in this case. I'm bringing only one of the examples, so you may jump immediately to more than one mobile beacon per one submap that you've built. It's one path, so you increase the number of mobile beacons. But we recommend that you go a different way. So instead of, for example, one submap in 2D, you make one submap but in 3D because eventually you need to build 3D submaps. That's great. So once again, remember, you do not change multiple things at the same time. You change only one thing. Inverse

3:56 Architecture is already there. One mobile beacon, one hedgehog is still there. What do you change? You change from 2D to 3D. Of course, it immediately brings instead of two stationary beacons, four stationary beacons. And of course, all of them must have a different ultrasound frequency. That's what you check. But the change is simple and single: from 2D to 3D. Okay, you checked. You achieved the perfect now 3D tracking. That's great. Then again, you have two options: either you can increase the number of mobile beacons in the same 3D submap, or since you eventually need a larger area with multiple 3D submaps, you move to several. In this case, two, because Inverse Architecture in general and 3D particularly is more complex, so you do

4:54 not jump to multiple, multiple, multiple submaps, but you build another submap. In case of 3D, it means that you already have one, two, three, four stationary beacons, and you build a 3D submap. In order to build another 3D submap, you build another, or you add another two stationary beacons. They also must have different frequencies from those four. So it means that you have two common—so it means that you have two submaps. One is, for example, 19, 31, 37, 25, and another submap. They will have two overlapping beacons: 19, 25, 45, and 22, for example. So you build those kinds of things, and now you build the service zones because you must build the service zones for two or more submaps. With a single submap, you kind of build—you can build without

5:52 defining the service zones explicitly. But there are always some results, even if you don't define them. The radius is 30 meters, so the service zones are there. But when you build multiple maps, you must define those service zones anyway, because otherwise, when the beacon, the mobile beacon in the hedgehog, is in this area, it must be tracked only by this submap. But if the maps are not too large, then it's possible that some map from this area will be able to attempt to track a beacon in this area, and this kind of minority report or minority opinion is not great because the system may be confused, particularly when you have many submaps and they will be overlapping. So in order to avoid that, you must define the service zones. But that's a subject to another video. Please check them about the submaps and how to build submaps. The whole point of this video is

6:51 increase one step at a time complexity. So from Non-Inverse Architecture and one submap and one mobile beacon, you move to Inverse Architecture, the same number of beacons, but from Non-Inverse to Inverse. Then from Inverse Architecture 2D, you move to 3D. It automatically requires increasing from two stationary beacons to four stationary beacons. And the last step that we just discussed, you increased from one submap of four beacons to two submaps of six beacons—four and four, and two overlapping beacons or common beacons for both. So as soon as you build this, then you can increase the complexity step by step basically by adding thoughts above force, et cetera, et cetera, as many submaps as you wish. Still, you have only one single mobile beacon, a single hedgehog. As soon as you build it, and you

7:49 basically cover the whole warehouse, you're ready to introduce, you know, more fish, so more hedgehogs, mobile beacons in your network. Introduce them, and then it will be basically the end. If your task was just many mobile beacons—like many forklifts, so many people—but in case you have a task to introduce other robots, then introduce the robots even on a later stage. So, for example, you have many people already running around. That's great. But then introduce the robot because again, each step brings some additional load, some additional uncertainty because you may be mixing up something or confusing something. So introduce one complexity at a time. Other path: let's say from basic Non-Inverse Architecture to

8:48 more complex than Inverse Architecture with several submaps and then, let's say two, eventually multiple submaps, multiple mobile beacons, and multiple stationary, still 2D. That would be your path. Then the same but Non-Inverse Architecture in 3D, and so it will be the same but one step less than comparing it to Inverse Architecture because nonetheless, Non-Inverse Architecture is simpler, simpler. Let me repeat why: because for Non-Inverse Architecture, you can install stationary beacons, for example, Super-Beacons, of any frequency and it will still work because remember, Super-Beacons can receive ultrasound frequency on any frequency but emit only on the native frequency of the Super-Beacon. So it means that you may have whatever random frequencies and another connection will still work. For Inverse Architecture, it will not work because

9:48 You must have a very specific frequency. They must not be at the same frequency in the same submap, etc. So this is why start with the Non-Inverse Architecture, and then only move to Inverse if needed. So I hope it's clear by now about the steps: one basic step at a time. Don't jump over. Don't jump over to the final configuration. Then, other recommendations. Why do we recommend Super-Beacons to start with? Because Super-Beacons are the simplest. Let's compare it with other beacons. For example, Mini-RX. Super-Beacon has a hemisphere, so 360 degrees reception diagram in reception and transmission, by the way, diagram in horizontal if the beacon is looking upward, and 180 degrees vertically. For Mini-RX, for example, that's not

10:47 the case, because Mini-RX doesn't emit anything. So it means that you cannot build, for example, the map. You must enter the distances. Super Mini-RX has around 120 degrees diagram, not 180 degrees diagram. Mini-RX has a smaller antenna, so in case you have large distances or there is interference, etc., etc., there's a high chance to have a problem due to this. So Super-Beacon is simpler. So start with something simpler. So this Super-Beacon, later on, of course, you will replace it instead of the mobile beacon. Instead of Super-Beacon, you will use a badge, for example, or a jacket or watch or whatever, which is more demanding in terms of coverage, in terms of everything. But start is something simple, which is Super-Beacon.

11:45 The same with Industrial beacons. So Industrial beacons, they don't have embedded battery, so you need to provide the battery. You need to provide the electricity. It's easier to mix up something, confuse something. So this way, we do not recommend to jump immediately to Industrial Super-Beacons, for example, or Industrial-RX. Start with Super-Beacons, achieve, and then replace if needed, like in the previous configuration. You sometimes even achieve the perfect tracking in basic multiple submaps map and multiple hedgehogs based on Super-Beacons, and only after that replace them with Industrial beacons, for example. Similar, we usually recommend when you go to a very complex and demanding site, like industrial site, for example.

12:43 There's noise, acoustic noise, and you're not sure. Of course, eventually, since it's noise, since it's polluted or, you know, dust or something, of course, eventually you want to have Industrial beacon there, Industrial Super-Beacon there, or Super-Beacons outdoor. That's fine, but don't jump there. Don't jump immediately to the final configuration because maybe it's so noisy that our system will simply not work. It's also possible, not very probable, but possible. And that recommendation is very simple: get the Marvelmind starter set, test it in the real environment. You know, put it there, and you will clearly see using the embedded oscilloscope whether it works or not. If it does, then okay, 95% of issues are resolved, and you can be confidently moving to, let's say, real configuration, which would include, for example, Industrial Super-Beacons.

13:43 You know, quickly deploy the Super-Beacons because they have batteries, and later on, when you have more time and more confidence, you deploy already using Industrial Super-Beacon because you need to, you know, bring the power. Okay, my cat came. The same about the modems. Super-Modems is, of course, great. Okay, animal. Super-Modems are great, but Super-Modems are more complex to implement. So logic is very, very, very the same. If you want to deploy a very large configuration, even basic, no, not basic, but modern version 5.9, 5.1 is good enough. But of course, Super-Modem brings Wi-Fi connectivity, brings UDP or Wi-Fi. It brings significantly more

14:41 capacity. That's great, but it comes also with the complexity, special connector, special power supply. You cannot just connect. Okay, the latest modems, Super-Modems, you can connect already USB and use it. But, for example, the previous version of Super-Modems, you need the external power supply for sure. Then this Wi-Fi, don't mess with those small settings now. Later on, at the final stage, when everything is clicking, of course, use the Super-Modem, and of course, this is just a small highlight because there are more steps like, you know, uploading the software, pressing the default buttons, or I'm not touching them in this video because it's well described in some of the previous videos like "Avoid Typical Mistakes," and of course, the operating manual. Operating manual is your key guide for this step-by-step. This video was focusing on one single thing: don't jump immediately to the final configuration. Start with the basic configuration, which is Non-Inverse Architecture, two stationary beacons, one mobile beacon, achieve perfect tracking, and then increase complexity one step at a time. Thank you very much.

Key Takeaways

  • Start with a simple 2D NIA configuration using one mobile beacon before expanding
  • Achieve perfect tracking accuracy at each complexity level before adding more beacons or coverage areas
  • Incremental expansion prevents debugging nightmares and accelerates time-to-value for autonomous robots and warehouse automation
  • Test each new component independently to isolate issues and maintain system reliability
  • Rushing to complex configurations sacrifices positioning accuracy and extends commissioning timelines
  • Validate line-of-sight, antenna placement, and radio setup during baseline testing to support future scaling

👥 Relevant For: Engineers & System Designers

Robotics engineers, warehouse automation managers, and integrators deploying indoor positioning systems (RTLS) who need to establish accurate multi-beacon tracking environments. This guide solves the complexity problem of expanding from simple 2D positioning to advanced warehouse-scale indoor navigation without sacrificing positioning accuracy.

? FAQ

Q: What is the optimal starting configuration for an indoor positioning system?
Begin with a simple 2D NIA setup using a single mobile beacon. This baseline establishes whether your infrastructure (antenna placement, line-of-sight, radio setup) supports accurate tracking before introducing additional complexity.
Q: When should I add more mobile beacons to my indoor positioning system?
Only after achieving perfect tracking performance with your current configuration. Adding beacons prematurely prevents you from identifying whether issues stem from hardware setup, software calibration, or the new beacon itself.
Q: How do I validate tracking accuracy during system expansion?
Test each new component (additional beacon, coverage area, or floor) independently with controlled movements. Document accuracy metrics before proceeding to the next complexity level. This prevents cascading failures.
Q: Can I skip steps when deploying an indoor positioning system for warehouse automation?
No. Even for mission-critical applications like forklift tracking or autonomous robots, incremental validation ensures reliability. Shortcuts typically result in longer troubleshooting and reduced system performance.
Q: What common mistakes occur when building complex maps too quickly?
Organizations often deploy multi-beacon systems across large areas without baseline validation, making it impossible to isolate root causes of positioning errors. Rushing this process delays production use and increases total deployment costs.

Technical Background & System Details

Building robust indoor positioning maps requires a methodical, incremental approach—and this guide explains exactly why. Many organizations fail to achieve reliable indoor location tracking by attempting overly complex configurations immediately. Instead, successful deployments follow a proven progression: begin with a basic 2D NIA (Non-Intrusive Array) setup using a single mobile beacon, verify flawless tracking performance, then systematically add complexity. This strategy applies across all applications: autonomous indoor robots, drone navigation, forklift tracking, and warehouse automation systems. Each expansion step—additional beacons, multi-floor environments, or extended coverage areas—must be validated thoroughly before proceeding. The guide emphasizes that rushing this process compromises positioning accuracy and creates debugging nightmares. By following this incremental methodology, integrators achieve stable RTLS performance, reduce commissioning time, and establish a solid foundation for scaling indoor positioning systems throughout facilities. Perfect tracking at baseline configurations ensures that subsequent enhancements maintain system reliability and performance.

📍 Need precise indoor positioning for your project?

Plan Your System →

For more information, please contact: info@marvelmind.com

Scroll to Top