Mobility & Rail covers connected transport: fleet telematics, ADAS, in-cab driver monitoring, sensor fusion, LiDAR-based railway track safety and always-connected asset tracking. These are environments where the sensor is dirty, the power is intermittent, the network drops for hours, and a false alarm costs as much as a missed detection.
A rail corridor is long, mostly empty, and occasionally occupied by something that must be detected before a train arrives. Camera-only approaches struggle at night, in fog, and against low-contrast obstructions. LiDAR gives range and geometry independent of illumination, which is why it anchors the sensing.
The hard part is not detection. It is the false alarm rate. A system that stops trains for foliage movement will be switched off within a fortnight, so classification, temporal persistence and zone geometry matter more than raw sensitivity. Fusion with camera imagery is what makes a confident classification possible.
Vibration, temperature, ingress, power availability, mounting constraints and network coverage measured rather than assumed.
Field of view, range and geometry established against the actual site, with occlusion mapped.
Real operational data captured across conditions, including the rare events the system exists to catch.
Detection, classification and alarm logic developed with false-alarm cost weighted explicitly.
A limited deployment run long enough to see seasonal and diurnal variation.
Provisioning, over-the-air update, monitoring and support at scale.
Faststream works on LiDAR-based track and corridor monitoring, combining LiDAR range and geometry with camera imagery and edge inference to detect obstructions and intrusion. The engineering emphasis is on classification and false-alarm rate, because a system that alarms on foliage will not stay switched on.
Yes. In-cab fatigue and distraction detection using infrared imaging, engineered to work with the driver wearing glasses, at night, and across a wide range of cab geometries and seating positions.
Data is buffered on the device and forwarded when coverage returns, and where continuity is required the device falls back to a non-terrestrial satellite link. Cellular remains the normal path; satellite is the exception layer.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.