Indoor Navigation 101 - An introduction to Indoor Navigation and Mobile Wayfinding

Mobile Wayfinding a.k.a. Indoor Navigation is hot. However, when you start searching on the internet, you notice quickly how different the images are that people perceive when discussing indoor navigation. And don’t even mention the huge mountain of terminology you are confronted with such as ‘beacons, blue dot, SLAM, inertial sensors, BTLE, routing, triangulation, dead-reckoning’... Are you still following? We have spelled it all out in our blog ‘Mobile Wayfinding 101’.

The two most important components that you need to differentiate between are:

  • Indoor Positioning
  • Indoor Navigation

Indoor positioning is simply determining the location of the cellphone of the user who moves into a building. Outside that is very easy. Outdoors your location can be determined by GPS satellites with an accuracy of about 10 -2 0 meters. Inside buildings, however, GPS does not work, so we have to find other ways to determine the location of a cell phone

Determining a location inside a building. How can you do that?

There are several ways in which you can determine the location of a cellphone inside a building. 

The most important distinction in indoor positioning is:

  • Geo-fencing / Proximity
  • Sub-meter-positioning

Geo fencing is the virtual delimitation of an area within a building. That can be done for instance with a beacon. A beacon sends a signal (blue tooth signal) and the phone can pick up this signal at a distance of 10-75 meters. This distance depends on how strong you set the signal strength of the beacon. If you set this to a radius of 25 meters, it means that you would fall within a diameter of 50 meters (2 x 25 m) within the virtual delimitation of this beacon. So it is not an exact determination of location, but that is probably not necessary anyway. More about this later.

Sub-meter positioning is determination of location that is much more accurate. An average accuracy of 2-7 meters can be achieved. Some claim an even higher accuracy, but that is nonsense and not even relevant. Following are a number of indoor positioning methodologies. There are tens of those, but most can be classified in the following categories or are a hybrid form of those.
 

Sub meter positioning with beacons

Instead of using just 1 beacon for an accuracy of approximately 50 meters, you can also install 3 x as many beacons, to realize a more accurate positioning of the cellphone. 
The phone receives a radio signal from 3 beacons and can measure the strength of these 3 signals (RSSI). The signal strength has a more or less linear relation to the distance of the beacons, so with some calculations the location of the cellphone can be determined.

Radio signals are messed up
In a lab or an ideal building with just four exterior walls, this will perfectly work. However, the accuracy of the signal is strongly influenced by objects in open areas, such as inner walls or pillars. These objects reflect and absorb the signal. But objects without a set location cause even more problems, such as doors, furniture, people, etc. They cause an unpredictable interruption of the signal, which makes it very difficult to receive true radio signals and therefor to determine the location. 
 

Sub-meter positioning with Wi-Fi stations

The same applies for Wi-Fi stations. The only addition is that the radio signals that are sent from here are also not constant. That makes location determination using Wi-Fi stations even more complicated. Devices running on iOS 4.3 and higher will not be able to provide client based positioning via WiFi. That means that almost the total population of iPhones will be excluded.

However, server side positioning will work up to 10 meter accuracy for all devices. In that case the building needs acquire additional hardware that is very expensive. For example Cisco MSE, includes a location server that can provide location data to smartphones.

Sensors in the phone

Every smartphone contains sensors, such as a barometer, a compass, an accelerometer, a camera, etc. These sensors can be used when determining the location of a phone in the area indoors. Data originating from these sensors is often used in combination with signals from beacons or Wi-Fi stations and is therefore referred to as fusion technology: a mixture of several elements that have to lead to accurate indoor positioning. Objects in the inner space often disrupt indoor positioning by for example beacons. The detrimental effects of those disruptions can be resolved by using for instance also the information from the accelerometer. The accelerometer indicates moving behavior of the user. So, with some physics and mathematics you can still easily determine the most probably position if the user starts to move and the beacon signal is disrupted, exactly at that moment.


Indoor positioning using ambient magnetic fields

This is a system that has been developed by a university in Finland. This system has now been commercialized in the form of the company IndoorAtlas. Every building has its own magnetic 'fingerprint'. IndoorAtlas has developed a tool with which you can create this magnetic fingerprint of every building and then the compass (together with some other sensors) can determine the location within the building. The huge advantage of this system is that you do not need to place external positioning hardware inside the building.


SLAM (Simultaneous localization and mapping)

This is a term from robotics and artificial intelligence. Robotic mapping is a way to have an autonomous system construct a map or floor plan (finger printing) from all kinds of signals (e.g. blue tooth beacons, Wi-Fi, camera, etc.) and localize itself at the same time.  Currently there are too many practical problems such as computational power, to have SLAM available and bring value to mobile indoor wayfinding.


What's the use of indoor positioning?

In principle, with just the location information, you can do nothing that has any value. Providing indoor positioning for your building is not an objective unto itself. Location determination is a means to an end. But what would be the objectives you have in mind?

Indoor positioning is embraced by the commerce as being the next big thing. With indoor positioning it is possible to send messages to people or to filter information based on their indoor location. The question only is: How exact does your indoor positioning have to be for the content to be filtered relevantly? Is it actually necessary to pursue a sub-meter positioning? It depends on the application you choose that utilizes the location of the user. For marketing purposes there is no single use case and/or business case yet, that necessitates sub-meter positioning.

Another application that can utilize indoor positioning is indoor navigation.  The question here is, again, how exact your indoor position needs to be.  This, again, depends on which type of application you choose.

Indoor Navigation, an historic perspective

The Dutch information scientist Edsger Dijkstra came up with the routing algorithm in 1959. Eventually this algorithm was named after him. Shortest route calculations are usually executed using the Dijkstra algorithm. 


Route planner CD-ROM in 1992

This system is actually not a real navigation, because the only thing it does is calculate a route and show this in a list of instructions to the consumer. So it doesn’t provide insight into which instruction should be read at what moment. So the content is not being filtered by the context of the consumer; in this case, his actual location. Route planners only show a global image to the user regarding the route he or she has to drive or walk.


Philip CariN in 1995

CARiN stands for CAR information and Navigation. This was the first commercial turn-by-turn navigation system for the outdoor environment. It set a complete new standard and it is still valid today. Apple and Google now dominate this market for turn-by-turn outdoor navigation.


Indoor Navigation; the various concepts

What we see is that indoor navigation is still very much in its infancy. For the moment the outdoor navigation concept is just being copied to an indoor environment, because suddenly possibilities arise to utilize indoor positioning. (Scientific) research is still needed before we can determine with certainty which indoor navigation concept is going to provide the best consumer experience. 

There is one important aspect within the indoor navigation and that is the automatic routing. The automatic calculation of routes is a must. To enter all routes manually is an enormous job. Imagine a hospital with 150 destinations (policlinics, restaurant, shops, chapel, nursing stations, etc.). Your goal is not just to be taken from the main entrance to those 150 locations (150 routes). You also want to navigate from all 150 locations to all the other locations. That means a total of 150 x 149 = 22.350 routes. The manual defining of these routes is a lot of work. But, suppose that a P.O.I. relocates, or that a corridor has to be closed off or an elevator is out of order? Then you have to able to intervene fast and you can only do that if all routes have been calculated.

Several concepts claim to be indoor navigation instruments. Here we will list the most common concepts.
 

Route descriptions

This is the equivalent of the outdoor route CD Rom. The consumer receives a set of navigation instructions so these can make the correct decisions underway to reach his end destination. This is a method that is being utilized if there is no automatic positioning present in the building. The consumer has to be aware himself of where he is and will have to retrieve the correct instruction that relates with that position in order to be able to navigate. This navigation concept occurs with automatic route calculation, but also with hand-defined routes.
 

Maps with walking route

These are digital maps where the walking route has been marked on a map from start to finish. The consumer has to be very good at reading a map and has to be able to translate his own position and the 2D information to the real environment. The consumer has to understand where his position is and has to try and follow the marked route. This navigation concept occurs with automatic route calculation, but also with hand-defined routes.
 

Maps with a blue dot (blue-dot navigation)

This is the best known indoor navigation variation. The reason for this is that people think of maps and a blue dot when they think of navigation because that is what they are used to in outdoor navigation. Many startups worldwide try to imitate outdoor navigation with navigation inside buildings.

For so-called 'blue dot' navigation you always need sub-meter location determination. It has to be visible where you are at all times. Continuous location determination is therefore needed during the entire navigation process. This demands the installation of an indoor positioning system that can provide sub-meter location determination.

This variety of indoor navigation we can also call turn-by-turn navigation. At every decision point the consumer receives a navigation instruction. In this case too, the user has to relay 2D information into the real environment. This navigation concept occurs with automatic route calculation, but also with hand-defined routes.
 

Photo landmark navigation

Photo landmark navigation is a set of navigation instructions whereby the consumer sees pictures of the environment with the correct navigation instruction in words and images. The user can see exactly on which location the indicated navigation instructions have to be followed.  You only have to compare the picture on the phone with what you see around you.
For this system, external location determination is not necessary because the user himself does this without realizing this. External location determination can be utilized to warn users when they deviate from the correct route so that the user can have a new route calculated fast and be presented with it. This navigation concept occurs with automatic route calculation, but also with hand-defined routes.


Indoor Navigation and Wayfinding

An indoor navigation system does not ensure that you can easily find your way inside a building. Indoor navigation is a concept which makes mobile wayfinding possible. First we have to determine a definition for 'wayfinding'.  Wayfinding encompasses all of the ways in which people (and animals) orient themselves in a physical space and navigate from place to place (Wikipedia). 

Wayfinding design seeks to provide structure and organization to build environments by providing the appropriate tools to enable self-guidance. It seeks to reduce complexity in spatial problem solving as people work their way through a given environment. 

There is a lot more to it than just the use and the combining of the techniques for indoor positioning and routing. It takes a lot more to let people actually find their way inside large buildings. Therefore, indoor navigation does not guarantee good wayfinding.  Technics is never the starting point. It’s all about consumer experience. That has to be perfect and technology should only fulfill a helping role with that.