Augmented Reality (AR) is a rapidly expanding market, with demand ranging from entertainment apps and games to custom business solutions.
In this new wave of AR solutions, AR indoor navigation is in high demand. However, the current state of technology is far from perfect. The best known indoor navigation solutions use beacon technology and only leverage AR for visualizing routes.
But AR can also be used for gaining more accurate user positioning data. Currently, robust mobile app solutions feature maps and built-in GPS to determine outdoor positioning and navigation.
The majority of smartphone owners regularly use navigation apps every day as part of their routine. Yet the situation is far less developed and more complicated when it comes to indoor navigation, such as finding what floor a user is positioned on.
Why you should Invest in Augmented Reality for Indoor Navigation
- Visual positioning systems (VPS) are better suited for larger areas
- Wi-Fi positioning systems can be expensive to set up and maintain
- Locations all have different accuracy thresholds
- AR visual markers can accurately map out smaller areas
- Wi-Fi RTT solutions are more accurate than Wi-Fi protected setups (WPS)
Visual Positioning Systems Only Work Over Large Areas
GPS is a fantastic technology, but it falls short when trying to determine position within buildings.
While GPS technology is fairly precise on large premises like warehouses or airports, it cannot pinpoint which floor a user is on. This means that users need to select their floor manually when navigating indoors. This is how current apps like Google Maps or Apple handle the problem, but it’s far from ideal.
Google is attempting to resolve its shortcomings with indoor position by using a visual positioning system (VPS), yet it’s unclear as to when VPS will be ready for a full-scale deployment.
Apple has been working to solve the issue by augmenting ARKit with a new class called ARWorldMap, which creates feature points in the vicinity of the user in order to determine their position.
While ARWorldMap is already useful in certain applications, it’s lacking in many contexts for indoor navigation. Many offices and corporate spaces tend to have uniform hallways and rooms from floor to floor. Similar environments like those can confuse ARWorldMap. In addition, ARWorldMap can require intensive calculations, which can slow down function on most user devices.
Wi-Fi Positioning Systems Can Be Costly
Apple has created a navigation solution for more than thirty airports, that relies on a Wi-Fi positioning system (WPS). It uses received signal strength indication measurements from a variety of Wi-Fi access points to create a series of Wi-Fi footprints corresponding to the building.
This solution is able to offer location accuracy at around 5 to 15 meters and is dependent on certain factors like the congestion of walls, ceilings, floors, and human bodies, as well as how many networks nearby are available.
Beacon technology has been hailed as immensely promising for indoor navigation. In the Gatwick airport in England, 2,000 beacons were installed, and AR is used to draw virtual routes. Beacons are also great for proximity marketing.
But for indoor navigation, this solution has significant downsides. With unit prices ranging from $10–$20 per beacon and the need for biennial battery replacement, the upfront, and maintenance costs are far from negligible. Additionally, these devices have a working range of around 10 to 100 meters and have questionable accuracy.
Different Locations Have Different Accuracy Thresholds
The question is what’s the necessary accuracy threshold for a user to navigate indoors. Evidence points toward a threshold for minimum accuracy of no greater than around 10% of the minimum distance between any two destination points.
To illustrate, take a look at this visual below to get a better sense of how the accuracy threshold depends on the total distance between points of interest.
The image shows the minimum distance between terminals within an airport to be 40 meters. So, the accuracy threshold stands at 4 meters, which is 10% of the total distance. In offices or campuses where there are 4 meters between room doors, the threshold would be 40 cm.
With iBeacon technology, Apple states that beacons’ signal strength should not be relied upon for calculating distance values manually and that beacons are only able to give us an approximate distance.
Our internal testing has shown that beacons are able to deliver around 5 to 6 meters of accuracy after applying complicated algorithms. This isn’t accurate enough for wayfinding in a campus or shopping mall. And we are still faced with the problem of knowing which floor the user is on.
AR Visual Markers Are More Accurate at Smaller Scales
AR visual markers are an alternative solution that’s capable of positioning a user, bringing down the accuracy threshold down to within millimeters.
The markers are images that are recognized by AR software development kits (SDKs) like ARKit and ARCore. When a user scans a marker, their device transfers data about their exact location to notify where the SDK should position AR content.
Check out this diagram from our video on ARKit.
By placing a visual marker on a wall or on the floor and then scanning it, we get precise coordinates within the real world.
A possible app architecture is to be modular—consisting of a mapping module, positioning module, and a rendering module. Right now, mapping and rendering have been exhaustively studied, and it’s possible to upgrade and customize the modules to individual needs. Positioning is more complicated.
In order to navigate, we require a map. We can’t always rely on being able to find a map of a given building with the necessary flexibility, and we also run into security issues with locations like airports.
We create a floor plan after measuring the building with laser distance measurers, measuring tapes, and other traditional tools to make sure our map is 100% accurate. After doing so, we can implement visual markers for all relevant points, like elevators, hallways, doors and so on, placing them on 50-meter intervals.
The map and intervals are detailed in the following image.
The map shows a navigation graph along with AR visual markers placed at regular intervals. If we develop an indoor navigation app for a small building, we need about 15-20 visual markers.
But if an indoor navigation solution for a supply chain in various areas is being developed, we have to deal with hundreds of visual markers. The performance may decrease since every one of those markers should be checked for matching.
With the help of GPS, we can pinpoint the user’s current location. Then, we only need to work with the visual markers in that building.
We can build dynamically updated maps giving more benefits to venue owners. For instance, let’s take a grocery store with a deli section in the back right corner. If the section shifts to the back left instead, the app updates the position and is able to generate a new route to the new location.
Apple has begun the Indoor Mapping Program, allowing owners of businesses to build standardized, detailed indoor maps of their locations. This may have future utility with AR indoor navigation.
Wi-Fi RTT Is More Accurate Than Wi-Fi Protected Setup (WPS)
Although utilizing AR for user positioning offers a host of improvements over beacon-based wayfinding, it’s still necessary to solve several issues before AR indoor navigation is the clear-cut choice.
And the requirement for an uninterrupted session is the biggest hurdle to overcome. It’s necessary for the user to maintain an active camera from the first marker all the way to the final destination to hit the accuracy threshold.
A way to get around this constraint is to use Wi-Fi RTT-based positioning, which potentially has far greater precision than currently used WPS solutions.
Rather than assessing signal strength, the Wi-Fi router calculates how long it takes for a signal to go to the user’s device and then how long it takes for a signal acknowledging receipt to bounce back. This is known as Round-Trip Time (RTT).
The real power comes in when there are over three Wi-Fi Routers in an area, which allows for an algorithm, based on the principle of multi-lateration, to achieve accuracy in the range of 1 to 2 meters.
As of August 2018, Wi-Fi RTT only has support at a platform level from Android Pie. This means that a given smartphone doesn’t have to be connected to a Wi-Fi router for the system to function — the smartphone itself determines distance values.
It’s likely that other tech companies will follow suit with Wi-Fi RTT, and this will have great significance in the fields of location-based marketing, home automation, and any indoor navigation-based field.
Augmented Reality Shows Promising Results for Indoor Navigation
Many of the supporting technologies involved with indoor navigation aren’t fully mature at this point, but they’re already producing impressive results.
As more companies explore the potential of AR visual markers and Wi-Fi RTT sees greater adoption and proliferation, AR-based indoor navigation will continue to progress.