Geolocalization is key to multiple applications and services. Indoor Positioning Systems (IPS) are key to extend these services indoors where GPS systems are not reliable. IPS are integrated and can be deployed in homes and buildings, and can interact with smart devices to provide them with spatial context. In this post we will explain how IPS work, what types of IPS technology exist, its features and its main applications to track people, from hospitals to market research.
What is an Indoor Positioning System?
An Indoor Positioning System (IPS) is a system that is able to locate one or more people and objects in an indoor environment. An IPS is usually composed of two different elements: Anchors and location Tags.
Anchors are devices placed in the building, while a tag is carried by the person or object whose location is of interest.
Figure 1. Indoor positioning illustration.
Why do we need Indoor Positioning Systems?
Real-time knowledge of the location of people or objects has become essential for the deployment of services in many fields such as retail, logistics, urban planning, leisure activities, etc. The success of global positioning systems and the mobile revolution have forever shifted the way we relate to technology both in business and in our personal lives.
However, GPS is not always able to locate people when they are inside a building. This is because GPS technology uses the signals of satellites in orbit. These signals are seriously degraded when there is no direct visibility, which makes indoor location finding difficult.
Indoor navigation and location systems have appeared to fill this role to provide location information in real-time in indoor environments.
Types of Indoor Positioning Systems in terms of technology
There are different indoor tracking systems depending on the type of signals used (see Figure 2). All of them have their own limitations and none of the options is perfect. From a practical point of view, it is important to understand them to select, depending on the application, the best technology. Below, we explain some of the most common technologies.
Figure 2. Main Positioning Technologies
IMU (Inertial Measurement Unit)
Inertial Systems inform about the relative movement of the tag with the integration of several sensors such as accelerometer, magnetometer, and gyroscope, in a tiny module. These sensors are useful in determining the direction and orientation of movement. Combined, they can provide an estimate of the relative motion with regards to the previous position. This information is usually obtained by combining all available signals using algorithms such as Dead Reckoning. Dead reckoning is “the process of estimating known current position based on the last position and incrementing that position based on known or estimated speeds over elapsed time.” (Farid et al.,2013)
One of the advantages of this technology is that it does not require the use of anchors in the environment. Unfortunately, the accuracy of this type of systems is usually poor, as the error accumulates over time and can be on the order of meters after just a few seconds.
For that reason, it is common to use it in combination with other technologies to smooth the results and reject outliers. Also, the ability to detect movement can be useful to detect if the tag, and therefore the participant, has stopped.
These systems used infrared light pulses like a TV remote control. They require an unobstructed Line of Sight (LOS) between the anchor and the tag.
This type of system can be used as a very reliable room detector. Since light cannot traverse walls, it is not possible for a tag to detect light from an anchor without being in the same room. For precise localization, they require installing many anchors and can struggle due to the low quality of the signal strength measurements required to compute the position from multiple anchors. A similar approach is currently used in VR headsets to accurately locate the person in a room by using multiple light sources and reflecting objects.
Figure 3. Infrared behavior with obstacles. Original can be found on this webpage
Ultrasound systems use sound instead of light. “Ultrasound wave is a mechanical wave that is an oscillation of pressure transmitted through a medium.” (Al-Ammar et al.,2014). It does not interfere with electromagnetic waves and does not require line of sight. The system requires a set of anchors and a tag. It uses Time of Flight, that is, the time required by sound to travel from an anchor to a tag or vice versa, to estimate the distance between them. Once you have at least three distances available, the position can be computed using trilateration (see Figure 4).
Figure 4.Trilateration: the basic concept is to find the intersection of the three circles defined by the three distances between the anchors and the Tag.
Ultrasound systems are not the most common in applications. They require the placement of multiple anchors and time synchronization between bluetooth beacons (anchors)anchors and tags. They can achieve submeter accuracy. However, ultrasound signals are affected by interference from solid objects and consequently accuracy can be poor if these are not considered.
Radio Frequency Technologies
The most common IPS are based on radio frequency signals. There are two reasons for this. First, some systems reuse technology that is already deployed, including WIFI networks, Bluetooth, and mobile phones. This drastically reduces the cost of deployment and makes the technology easily accessible to a larger number of people (e.g. by installing an app).
The second one is that, as these signals can traverse obstacles, they can work in real-world settings where obstacles are unavoidable, including commercial areas. In some cases these systems can also provide larger coverage. Below, we describe the most used IPS systems using radio frequency:
RFID (Radio Frequency Identification)
RFID based systems fall in the category of “non-contact automatic identification technology” (Song, et al.,2011). RFID uses electromagnetic fields to identify and track tags attached to people or objects. This type of system uses RFID tags and readers. Readers send pulses that are detected by tags. Tags respond to the request of a reader by sending back a small amount of information, such as an ID.
These systems send and receive radio frequency signals from 125KHz (short-range systems) to 5.8GHz (long-range systems), although they normally use ISM bands. RFID can be passive, active, or semi-passive (Parkash, et al., 2012). The simplest RFID systems use passive tags that obtain the power necessary to send the answer directly from the reader’s pulse. Passive tags are very cheap, can only store a few kb of memory, and the reader has to be within around 1 m of the tag to get the information.
Figure 5. RFID passive. Taken from: https://www.indiamart.com/
Semi-Passive RFIDs have a battery to power the tag and other simple functions, but the antenna and overall functionality is similar to the passive RFID. These can provide an extended working range. Active RFIDs have different antennas so as to provide a larger range, around 100m. These tags can also store more information. However, they cost more and are often larger..
As mentioned above, RFID systems are good options to identify and detect the presence of people and objects. They do not require line of sight like infrared systems. However, they do not provide tracking information, and many times they are combined with other technologies that can provide positioning.
However, they may suffer from interference in environments with liquids, metal, or other sources of radio interference.
WIFI & Bluetooth
Some IPS systems use well-known Bluetooth & WIFI technologies. The main advantage of these systems is that they can use the pre-existing network infrastructure and that both WI FI and Bluetooth are available in mobile phones and other wearable devices. This makes them easy to deploy and cheaper than ad-hoc installations.
The main operating principle consists of using the Received Signal Strength (RSS). The strength of the signal depends on the distance between the sender and the receiver (see Figure 6). By simply measuring the RSS of the tag (e.g. a mobile phone) to multiple WI FI access points or Bluetooth beacons (which act as anchors), it is possible to estimate the position of the mobile phone using trilateration, the same principle used in ultrasound IPS.
The main difficulty for these systems is that WIFI and Bluetooth signals vary enormously in the presence of obstacles and moving people. Also, different materials affect the signals differently which affects accuracy. To overcome this, some IPS create a map of RSS specific for a given area based on ad-hoc calibrations (see Figure 7). The accuracy obtained with this type of systems can reach 1-2m.
Figure 6. Exponential drop of received power (RSS) with distance (meters). When moving far from 2 meters, there is no difference in RSS so the accuracy in the distance measure is poor. Original can be found on this webpage.
Figure 7. Map of RSS with three Anchors. Areas near Anchors are represented by green colour. Reading signals from 3 anchors and comparing them with signals previously recorded in each cell of the map, position can be estimated. Original can be found on this webpage.
Although they share the underlying principle, there are some differences between Bluetooth and WIFI. On one hand, WIFI offers great coverage, although power use is considerable and something to be taken into account. On the other hand, Bluetooth requires less power, especially Bluetooth low Energy (BLE) technology. Nevertheless, reducing power means reducing coverage too. Bluetooth 4.0 has an ideal maximum range of 100m with high data rate (up to 2.1 Mbps) while BLE only has around 60 m with LOS and with a substantial reduction in data rate(125kbps) . ZigBee is another less-known wireless communication protocol that operates in the 2.4GHz band and that is also used in IPS systems.
An alternative radiofrequency technology to WIFI and Bluetooth are ultra-wideband (UWB) based IPS. “UWB is a radio technology for short‐range, high‐bandwidth communication holding the properties of strong multipath resistance and to some extent penetrability for building material which can be favorable for indoor distance estimation, localization, and tracking.” (Mautz,2012)
The main advantage of this technology is the ability to penetrate materials such as concrete, glass, and wood, which makes it appropriate in typical indoor environments where line of sight is often not possible. Besides, the larger bandwidth means a high time resolution. This allows the measurement of time of flight between the sender and receiver, which results in better distance estimation that those obtained with RSS. UWB systems, therefore, use trilateration (Figure 4) to estimate the position of the tag from the distances to a set of at least three anchors deployed in the environment. The accuracy of UWB is currently the best of IPS with errors on the order of 30-50cm.
There are some disadvantages for UBW IPS. First, the technology is not readily available in the buildings or mobile phones of users and it requires an ad-hoc deployment. Since it occupies a large frequency bandwidth, there are relevant legal restrictions in order to prevent interferences between other radio frequency signals: 1) the permitted frequency band goes from 3.1GHz to 10.6GHz and 2) signal power is limited, which constraints the operation range to 100m or even less if data load transmission is required in the system.
Important features of IPS
To determine which technology is suitable to build your own IPS, some features have to be taken into account and balanced for the required specifications of the project.
- Accuracy: Defined as “the average Euclidean distance between the estimated position and the true position.” (Al-Ammar et al.,2014). Accuracy is considered the main feature in most indoor mapping systems and the most challenging to improve. The best solutions usually require ad-hoc deployments, which increase cost and complexity. For that reason, when accurate positioning is not critical, cheaper, and simpler technologies are used. The following table provides the best accuracies that can typically be obtained with the systems described previously.
Table 1: Technologies Accuracy in meters.
- Coverage & Scalability: This is the second most important characteristic rivaling with accuracy. Coverage is the area where the location information is available. The coverage of IPS usually ranges from a room to scalable systems that can cover multi-room environments or large areas such as warehouses or commercial centers. Often there is a trade-off between coverage and accuracy, where technologies with larger coverage typically imply smaller accuracy (see Figure 5 for a comparison of the technologies discussed above).
Figure 8. Technologies Accuracy vs Coverage
When choosing a technology, it is important to consider its scalability, that is, the ability to cover larger areas by adding more anchors, access points or card readers. Another important point is the ability of the system to locate multiple objectives at the same time.
- Adaptiveness: Changes in the environment may affect the performance of the system. (Farid et al.,2013) Because of this, the ability to cope with these changes is essential when accuracy is required.
- Sampling Rate: The number of positions obtained per second is also a compromised feature. Higher rates usually require more complex systems that require more computational power and more energy.
- Cost: This includes deployment costs, operational costs, and maintenance during the lifetime of the system. Some technologies require fixed installations while others are mobile or can use the existing infrastructure. Those technologies using trilateration usually require calibration, which is costly in time, especially if the installation of the system is not permanent.
Real-world examples of IPS in use
IPS can be used in many different applications that use the position of moving objects or people indoors. Smart buildings, including hospitals, factories or warehouses, can use this information to improve security, efficiency, and automate operations. Commercial centers and brands can also use the position data to provide targeted promotions, to better understand the behavior of their customers and to optimize their spaces. Below, we show some of the most popular use cases:
The use of location systems in factories has been increasing during recent years, with the goal to avoid accidents and save time in processes. Accuracy is crucial in order to ensure a successful implementation.
- Airports & Railway Stations
To implement IPS, people download a specific application. With the goal of saving time, the app provides useful information such as how to arrive at the gate, or where to find an information booth.
Besides, owners can control the flow of people in some areas in order to improve the facility installations.
- Market research
Market research, especially using neuromarketing, is one of the main areas of application. The flow of people is analyzed in specific locations inside the shop in order to detect areas where customers do or do not pass by.
An advantage of this technology is that, when combined with other neuromarketing techniques such as EEG, GSR, BVP, or eye tracking (eye tracking devices), it enables registration of emotions felt by participants throughout the entire route, identify where exactly the emotions were felt and what has caused them.
The following video shows the trajectory of a participant in a real point-of-sale study obtained with Indoor Positioning System, as well as the impact and emotional activation, obtained with Ring; and visual attention, obtained with Tobii Pro Glasses 2.0.
The following video shows the heat map of the grouped results from the same neuromarketing study.
- Supermarkets & Shops
IPS are used to facilitate the shopping experience. Normally, a location-based application guides the shopper around the shop to find their desired products, informing about the stock and the existing offers when they pass near a product.
In line with the other applications, IPS is useful for guiding people inside the building. They are also useful to inform doctors about an emergency if they are near the room where it is happening. They are also used to locate patients within a certain range, and to notify doctors if someone is out of range.
Figure 10. Monitoring and asset tracking in the healthcare sector using Bluetooth Low Energy (BLE). Taken from this webpage.
About the author
Leyre Morlas. - Electronic Engineer at Bitbrain (LinkedIn)
Leyre Morlas obtained her degree in electronics and automation engineering (2015) at the University of Zaragoza in Spain. Since then she has worked as a research scientist. Her main research interest is the design and development of indoor locating systems centered on wearable products. For this project she is responsible for both hardware and software design.
- Zahid Farid, Z., Nordin, R., and Mahamod, I.(2013). Review article: Recent Advances in Wireless Indoor Localization Techniques and System. Journal of Computer Networks and Communications, 2013, 185138. https://doi.org/10.1155/2013/185138
- M. A. Al-Ammar et al. (2014). Comparative Survey of Indoor Positioning Technologies, Techniques, and Algorithms. International Conference on Cyberworlds, Santander, 2014, pp. 245-252. https://doi.org/10.1109/CW.2014.41
- Song Z., Jiang G., Huang C. (2011). A Survey on Indoor Positioning Technologies. In: Zhou Q. (eds) Theoretical and Mathematical Foundations of Computer Science. ICTMF 2011. Communications in Computer and Information Science, vol 164. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-24999-0_28
- Mautz, R. (2012). Indoor Positioning Technologies, Habilitation Thesis at ETH Zurich. https://doi.org/10.3929/ethz-a-007313554
- Parkash D., Kundu T., Kaur P.(2012). The RFID technology and its applications: A review. Int. J. Electron. Commun. Instrum. Eng. Res. Dev., vol. 2, no. 3, pp. 109–120, 2012. https://www.researchgate.net/publication/232575248_THE_RFID_TECHNOLOGY_AND_ITS_APPLICATIONS_A_REVIEW
You might also be interested in:
- What is an Implicit Association Test (IAT)?
- Introduction to Implicit Bias Tests: IAT and Priming implicit tests
- All you need to know about neuromarketing
- Examples of companies using neuromarketing
- Top 10 sources of neuromarketing news and blogs in Spanish and English
- 7 journals to consult for neuromarketing research papers
- Ethics and neuromarketing implications for market research and business practice
- 3 neuromarketing TED talk videos
- 20+ popular heuristics and cognitive biases
- The decision making process: Do we make rational decisions?