This year will be the 13th anniversary of the Viennese LBS Conference, an event which unites enthusiasts and experts of all kinds of Location Based Services, covering wayfinding, smart cities, visualization for LBS, geo-tagged Big Data and much more. Our research team is once again out on their mission to present their latest work around Indoor Positioning there, on November 15th. Get all the details and reserve a meeting now!
Indoor Navigation Know How at your hands
Have you ever heard of the FLIP? And no, we’re not talking about the gymnastics exercise (although our researchers are quite sporty, we have never seen any of them practising flips, but who knows what might happen behind closed doors?)
The FLIP is rather a tool developed in our research laboratory and stands for “Flexible Indoor Position Estimator”. It is designed for Fingerprint-based positioning systems and impresses with an excellent flexibility to be used for different application needs and hardware capacities.
- FLIP map
Minimal representation of radio map optimized for speed
- Hierarchical execution model
First finds building and floor, then general area, then the exact location
- Accurate uncertainty estimation
The size of the blue dot is a required input for Kalman Filters and advanced analytics
Figure 1: FLIP Map Figure 2: Comparing metrics between Reference and Observation
In FLIP, the radio fingerprint point references are transformed to a distance metric per visible point for each transmitter. The metrics are used to measure the observation-reference similarity, that in turn is used for coarse selection and as weights in the location estimation.
Learn from the best
While FLIP estimates an absolute position, we can improve accuracy using it in a Kalman Filter that includes a motion model. To estimate the motion, we use the accelerometer, gyroscope and magnetometer integrated in the smartphone. For this to work, we must first correct for the pose of the phone, which is done in an orientation filter. Affiliate Researcher Andreas Ettlinger is presenting our approach to this during the Positioning I session (see below).
Intrigued? Join the event and find out all about these fascinating discoveries at first-hand from our researchers.
All in all, we’ll be actively participating and presenting on Tuesday, November 15th, starting with our indoo.rs introduction by Boxian Dong to the paper presentations and interactve session. We’d also like to listen to your ideas or questions on the other days of the event, so if you spot us, definitely say hi!
Here’s our schedule at the event:
Andreas Ettlinger, Hans-Berndt Neuner and Thomas Burgess
“Smartphone sensor based orientation determination for indoor navigation”
Thomas Burgess and Reinhard Müllner, presented by Boxian Dong (indoo.rs)
“indoo.rs: what we do, and recent improvements in radio positioning”
Reinhard Müllner, Thomas Burgess and Heinrich Schmitzberger
“FLIP – FLexible Indoor Position Estimator”
As always, we’d like to invite you to reserve a meeting prior to the event so we can both find the best time to have a chat on-site.
When: November 14-16, (See details in the agenda)
indoo.rs presentations: Tuesday, November 15th
Where: Vienna University of Technology, Main Building
Also! News update: indoo.rs publication
Our work on the orientation filter was published in “Progress in Location-Based Services 2016”, which you can find here. Search for the article on “Smartphone Sensor-Based Orientation Determination for Indoor-Navigation”!