Context Global Positioning Systems (GPS) available in many consumer products such as mobile phones has mostly solved the problem of navigation but remains a challenge for indoor locations. A possible solution exists with 802.11 wireless networks and Location Based Services (LBS) that are able to compute location of a Wireless Station (WS) using triangulation of telemetry such as Receiver Signal Strength Indicators (RSSI) from nearby Wireless Access Points (WAP). The WS coordinates have inaccuracies due to it being “a function of distance, geometry, and materials” (Mengual, Marbán & Eibe, 2010) making distance travelled calculation inaccurate. In an experiment plotting a moving workstation the estimated distanced travelled was 483 metres compared to the actual distance of 149 metres (322% difference). Content The LBS system receives data from the wireless network, computes location information for each workstation and stores the data for later retrieval. The data can be sourced from the LBS using an REST API that returns JSON formatted data. To enable comparison to the estimated calculation, a controlled experiment with a wireless station moving to 20 known locations and turning on the wireless interface for 90 sec periods at a time was conducted. The continual stream of coordinates from the LBS can change not only due to the WS physically moving but also due to the errors in the location calculation itself. These errors can be significant and render any distance calculation meaningless. The experiment captured the calculated position from the wireless network and the actual measured x and y coordinates of a workstation in 20 locations in an office building. The challenge is to figure out using the wireless location ways to improve the accuracy of the prediction. Field Name Description time - Conversion of Singapore time to to seconds, from 00:00:00 x - x axis coordinates of floor map in feet y - y axis coordinates of floor map, origin is top left of floor map cf - 95% confidence in feet of radius away from x and y client likely to be. realx - x axis measured coordinates of the real location of the test subject realy - x axis measured coordinates of the real location of the test subject Inspiration The distance travelled by the workstation was 149 mtrs. How close can you get to this calculation used the predicted locations ?