For autonomous robots, drones, forklifts, cranes, VR/AR, people, and other industrial applications
But with our little help and explanations, of course, citing Wikipedia and other sources, the terms become much clearer.
If some key terms are missing or even incorrectly explained, please email us at firstname.lastname@example.org, and we will address it.
This page explains the key terms used in indoor positioning and real-time locating systems.
Anchor – a common name for stationary beacons in ultra-wideband (UWB) systems. In Marvelmind Indoor Positioning System, anchors are called stationary beacons.
When an indoor positioning system (IPS) or a real-time locating system (RTLS) determines the location of a mobile object, it calculates it as a location of a mobile beacon (tag) installed on the mobile object in the coordinate system of anchors. Anchors (stationary beacons) are the reference points for the positioning of mobile beacons (tags).
BLE-based indoor positioning systems use AoA as a method to improve accuracy. Some reports claim that the accuracy improves up to 3 times and can reach ~1m, albeit for more expensive and complex BLE anchors.
Typical BLE-based indoor positioning systems have poor accuracy of 2-5 meters because BLE-based systems rely on RSSI to estimate the distance between beacons and tags. RSSI is an unreliable criterion since there is multi-path propagation indoors. As a result, at the same point with the same distance between the BLE anchor and the BLE tag, the radio signal strength (RSSI) may differ by 10:1.
Bluetooth with AoA allows combining two sources of data: 1) Distance estimation based on RSSI, and 2) Signal angle of arrival (AoA) based signal from several antennas placed on calculated distances. Combining the data and using sophisticated algorithms allows for decreasing the inaccuracy of BLE-based IPS systems.
AR/VR devices use different types of systems for positioning – inside-out and outside-in. Typically, for AR, it is an optical inside-out system.
Marvelmind Indoor “GPS” is an outside-in system. It is often used to complement AR systems with outside-in tracking to provide “ground truth” and help AR’s systems when the environment is too challenging for their embedded inside-out positioning system.
Stationary beacons are used as references (“GPS satellites”) for the mobile beacons (“GPS terminals”) in the indoor positioning systems. Of course, it is not a real GPS – it is GPS-like usage – “GPS”.
Mobile beacons are placed on the mobile objects and tracked by the system. Remember, the indoor positioning system typically doesn’t directly track mobile objects (robots, drones, forklifts, cranes, VR helmets, etc.). It tracks the mobile beacon installed on the mobile object.
Bluetooth Low Energy (BLE) is a wireless communication system. The BLE protocol is not designed for distance measurements. However, relying on RSSI for the distance estimate and employing sophisticated algorithms, it is possible to build decent indoor positioning and navigation systems particularly suitable for tracking mobile phones and guiding people with special navigation apps that use BLE-based positioning for navigation inside shopping malls, airports, and museums.
Typical accuracy of BLE is 2-5m at best, though it depends on the dencity of installation of BLE anchors. Closer anchors = better accuracy = more anchors = higher total infrastructure cost.
Using BLE with AoA improves accuracy up to ~3 times, and reports say that it reaches 1m, but it comes at the expense of more complex and expensive anchors.
Though, BLE is not designed for positioning but is still popular due to the possibility of tracking devices with embedded BLE tags, i.e., regular modern mobile phones. It differentiates BLE from UWB or ultrasound systems, which require tags given to people. In BLE, the mobile phone is a tag. Also, BLE tags and anchors have lower costs than UWB or ultrasound systems.
Although not entirely technically correct, we often use the terms BLE and Bluetooth as synonyms. The technologies are very close and can be supported by the same hardware, but they are not exactly the same and not even compatible.
Global navigation satellite system (GNSS) is used to determine the location of GNSS terminals globally as opposed to local positioning systems or indoor positioning systems. If GNSS worked indoors, indoor positioning systems wouldn’t be needed.
The majority of GNSS chipsets and terminals support several GNSS simultaneously. It increases the accuracy and reliability of tracking.
Global positioning system (GPS) is an American version of GNSS. The most known GNSS system is often used as a synonym for GNSS.
GPS works poorly indoors. Thus, there is a need for indoor positioning systems (IPS) or real-time locating systems (RTLS).
GPS typically gives around 5-10m accuracy with open skies. There is a flavor of GPS – RTK GPS – that gives cm-level accuracy.
But, taking into account the ease to use and overall “Apple flavor”, iBeacons became a synonym for beacons and indoor positioning technology at some point in time. Luckily, over the years, the flavor has faded away.
Indoor “GPS” is a brand name used by Marvelmind Robotics for our precise indoor positioning system.
Notice that it is Indoor “GPS” – not Indoor GPS because GPS doesn’t work indoors.
However, our indoor positioning system streams data in the NMEA format – native GPS format. Thus, your devices won’t even notice that they are connected to the Indoor “GPS” but not to real GPS. Thus, virtually no integration is required.
Inertial Measurement Unit (IMU) is an essential element of indoor positioning systems. Typical IMU for IPS is based on MEMS technology. Of course, there are other – significantly more accurate and hundreds and thousands of times more expensive IMUs, but they are rarely used for industrial IPS because of cost, size, power consumption and other characteristics that make their applicability for industrial applications currently impractical.
It is worth remembering that magnetometers work poorly and unreliably indoors due presence of magnetic field ferromagnetic materials and strong currents flowing around and distorting the magnetic field.
Thus, Marvelmind beacons don’t have magnetometers inside. But they have 6D MEMS IMU (3D accelerometer + 3D gyroscope).
IPS may exist without IMU, but IMU enriches it significantly with relatively minor additional investments – modem MEMS IMU is unbelievably capable. Thus, the users get along with location data updates, not only the coordinates but also acceleration, pose against gravity, and speed of rotation (gyroscope).
IMU/gyroscope doesn’t give the direction. It provides a change of direction. Thus, for direction, one has to either use a magnetometer/compass that doesn’t work well indoors and is not recommended, or use the Pair Beacons configuration that does provide the direction even in static, or rely only on the dynamic determination of direction – possible on with moving mobile objects and precise indoor positioning.
In the video below, the robot-car didn’t have the Paired Beacons but using IMU, precise Marvelmind IPS, and special algorithms, the robot was able to determine the direction and follow the waypoints properly.
INS can stand for Inertial Navigation System and Indoor Navigation System, which is confusing.
For the sake of clarity, let’s refer to INS as the Inertial Navigation System.
It shall be very clearly put in plain English. Purely inertial IPS for industrial applications doesn’t exist commercially. IMU-based indoor positioning systems are a misleading fantasia. There are so many articles around that propagate a different message that it is easy to start believing or question your sanity because it is hard to believe that so many people can be simply wrong. Copying incorrect information or omitting essential information doesn’t make it right. Thus, let me stress again: inertial indoor positioning systems for industrial applications or tracking on the phone based on the phone’s IMU do not exist today. It is technologically unachievable, unfortunately.
Why don’t IMU-based indoor positioning systems for industrial applications exist? The answer is straightforward: the drift of the accelerometer is too high to provide cm-level accuracy in any reasonable time. By reasonable time, I mean at least seconds or minutes. No, modern MEMS IMUs are incapable of that.
There are reports about space-grade, military-grade, navigation-grade IMUs, etc., that can extend the time of acceptable accuracy of INS to some 10-60 seconds, but for a practically unreasonable cost and still for a very impractically short time.
IPS with IMU sensor fusion exist, and it is one the most accurate, fastest, and most reliable implementation of IPS.
IMU sensor fusion approach achieves results that are unachievable with non-fused technologies by taking the best from both worlds: the IMU world and the world of another technology with which IMU is fused, for example, odometry or Marvelmind Indoor “GPS”.
But! INS with sensor fusion is not an INS anymore, of course. It is an IMU-based system with sensor fusion. The inertial part gives an essential input, but without other inputs – correcting imperfections of IMU, the INS simply wouldn’t work. Not it will be less accurate or less stable, etc. – it won’t work for practical applications at all. Thus, properly, call it a sensor fusion navigation system where IMU is just one of the sources of data.
Marvelmind Indoor Positioning System is used in a wide variety of applications. Thus, there are three architectures to serve the customers optimally:
IA is good for:
– People tracking
– Forklift tracking
– Multiple beacons without location update reduction
Marvelmind Indoor Positioning System can work indoors and outdoors equally well. The only difference is in ingress protection (IPxx) requirement toward network elements.
Remember: IPxx certifications says that the IPxx-level exposure doesn’t kill the device. It doesn’t require that the device can work during the exposure.
IPxx has so many fine prints that it is practically impossible to apply in real life without additional information. Though, IPxx certifications gives some ideas about ingress protection, but not more than ideas.
LIDARs are such a powerful tool for autonomous robots and indoor positioning for vehicles in general that it lures and misleads many users. Not trying to diminish the importance of LIDARs; practical notes:
– LIDARs actively used in SLAM
– SLAM is excellent for research but not so for real-life applications
– LIDARs are good for obstacle detection, with some exceptions
– LIDARs are NOT good for positioning in real-life (messy) factory
– High-grade LIDARs remains (and will remain) expensive
– Low-grade (1D or short-range) LIDARs became very viable. Use them
Thus, when the environment is simple, LIDARs work well for positioning. When the environment is complex and changing, LIDARs are “confused,” and proper RTLSs such as Marvelmind Indoor “GPS” or UWB-based indoor positioning systems rescue the situation.
– UWB doesn’t require LoS – not true! UWB requires LoS
The practical conclusion is very simple: if you want a precise IPS
Use Marvelmind’s internal radio, which is just another proprietary radio protocol comparable with LoRa, for data transmission to and from your mobile beacons:
The pap in the indoor positioning system is correct to call a map of stationary beacons.
The map is not a floorplan of your premises. You can put it as a background to the map, but the floorplan image is interesting to the end-user, mostly – not to the system. The system measures everything in the coordinate system of the map.