Participants appear in live visualization with 2nd read. E.g. if algorithm receives start detection with no delay, then markers show up at 1st decoder (or Track Box) to be placed 0.5 - 1 km from start.
In replay participants show up directly with start detection.
Automated recalculation of the replay when event is finished. Recalculation considers data the algorithm received with delay, also.
Our powerfull algorithm is constanly learning and improves the quality of its prediction with every event. For Predictive Tracking our system just needs to know:
Transponder ID of each participant you want to display (from data import),
Exact .gpx or .kml of the race course (so called "shadowtrack"),
Locations of timing hardware as decoders or Track Boxes
Hardware with own locations (e.g. GPS in Track Boxes) forward their positions to Racemap
Hardware without own locations: You define their locations at shadowtrack in Racemap.
Racemap processes data from your timing system into an amazing race visualization. With our AI-based features you apply Predictive Tracking efficiently and flexible.
Mapping within 30 m distance to shadowtrack:
If timing system is located within laps the system will be mapped on each lap automatically.
If system is located more than 30 m from the shadowtrack the prediction ignores its reads.
Moving timing system: Flexible location of timing system during the race, eg. place Track Boxes on cars or boats. Our algorithm considers the current location of every read for live prediction and replay.
Multiple contests detection: One timing system can detect several contests (shadowtracks) simultaneously. The algorithm assigns the reads to the correct shadowtrack, depending on transponder ID and contest.
High performance: algorithm calculates location & speed for > 50,000 participants simultanously, e.g. München Marathon
Filters for realistic visualization:
Extrapolation of average speed until algorithm receives read from next timing system location
Speed correction depending on difference between marker in map and real participant
As timing systems (especially Track Boxes) might miss detections the virtual marker moves on over the locations of following systems.
speed values of splits in shadowtrack improve prediction e.g. triathlons with different speed in transition and cycling.
Filter_max: reads with two times faster average speed than set speed of section are ignored.
Filter_min: reads with average speed of 0.1 x set speed of section are ignored.
To keep a minimum delay of forwarding reads from timing system to Racemap is from key importance for the real-time prediction.
How to place Timing Hardware? Think of strategic locations for your hardware along the race course.
1st & 2nd read approx 0,5 & 1 km behind the start.
Place hardware that way the algorithm receives data every 10 min for each transponder.
Events with more or less equal speed: place boxes equidistance.
Last read 1 to 0,5 km before finish.
Live Tracking of swimming events e.g. place Track Boxes on buoys. eg. https://racemap.com/player/wa-open-water-swimming-series-race-8_2019-01-25/
Transition area: T_in, T_out with decoders or with appropriate hardware like Loop Boxes
We strongly recommend to gain experiences at several small tests to learn how to work with the prediction and the hardware. We support your tracking tests, just get in touch.