Transport Mode Detection (TMD) means recognizing the mode of transport based on sensor gathered data. Detected modalities could include walking, cycling, driving, riding a metro or tram or taking a bus. Detection between different modalities, especially motorized and non-motorized, is fundamental for further analysis needs and requirements.
TMD has fascinated researchers for some time now. Main drivers have been the utilization of motion data in many applications as well as easy access to real-time motion sensor data produced by mobile phones. One major challenge in current GPS based approaches is the high battery consumption.
Our TMD approach uses efficient algorithms based on acceleration and deceleration profiles. Recognition is based on using commodity sensors, like accelerometer and gyroscope in current smart phones and wearable devices. This approach gives fast and reliable mode of detection with low power consumption.
We have also the possibility to add new modalities in our recognised transportation modes. For example detecting a car detection mode can be further divided by car models. Transport Mode Detection can be easily integrated to any application. It should also be able to work without data connectivity.
The motivation for Transport Mode Detection is the growing need in different kind of MaaS (Mobility as a Service) based services that requires reliable information in a near real-time about the recognized transportation mode in order to serve their own customers with the right product offering.
Insurance companies and ride sharing service providers are interested about the driving behaviour of their customers and drivers.