IoT Analytics

Mobility Pattern Analytics

Creating insights from mobility traces

Many devices connected to the Internet of Things can determine and report their own locations using GPS, RFID or Bluetooth beacons. Mobility pattern analytics enable insights from these types of mobility traces through learning and automatic identification of characteristic mobility patterns.

Object location data is acquired through sensors that either determine their own locations (e.g., GPS) or track other objects; both methods are sensitive to environmental influences that can lead to gaps and errors in object mobility traces. If applied analytics cannot effectively deal with these gaps and errors, wrong conclusions are drawn from data.

Understanding the movements of “things” and categorizing repeated movements / patterns of movement is a critical capability for applications that involve logistics and industrial production.

AGT_Web_MPA_illegal_taxi_for_mobility_sm

Use Case: Illegal Taxi Detection

Unlicensed taxis are a problem in many cities around the world. As evidence, London has a dedicated Cab Enforcement Unit responsible for enforcing the law relating to taxis and privately hired vehicles.

Using our mobility pattern analytics, we can analyze as many as 20 million license plate recognitions per day in real time. Our analytics cleanse the LPR data of false detections, and extract mobility profiles for every vehicle. A characteristic mobility pattern of regular taxis is learned (based on mobility traces), and based on those patterns, any vehicle with similar mobility traces is compared against an official taxi license database. Suspicious vehicles are automatically identified for further investigation.

Our algorithms are also used to analyze movement patterns of delivery vehicles, helping to identify efficiencies and inefficiencies, and the root cause of the latter. This is particularly relevant for supply chain optimization and fleet management.

 

Share this:
AGT