Developing the embedded sensor hardware that turns foot-floor contact into clinical movement data

Extracting meaningful skeletal alignment and movement data from foot-floor contact is technically demanding. The sensors had to capture position, orientation, speed, and acceleration from each step with enough precision and consistency to support clinical decisions, distinguishing genuine gait instability from normal variation, and detecting subtle changes in movement patterns over time. At the same scale, the floor sensors had to be thin and robust enough to be embedded in or placed on a floor surface without creating a trip hazard or degrading under continuous foot traffic.

Synchronisation between the sensor hardware and the application was a critical requirement. Movement analysis is only clinically useful if the data arriving from multiple sensors is time-aligned, a desynchronised reading introduces artefacts that could look like gait abnormalities when none exist. The embedded system had to coordinate data collection across both film sensors in the floor and wearable sensors on the patient, and deliver that data to the application in real time without latency that would compromise the analysis.

Power management added a further constraint. Sensors deployed in care environments, particularly those integrated into floor films across large areas, cannot be dependent on mains power at every point. The hardware had to support battery operation with low enough power consumption to make deployment practical, while maintaining the continuous measurement capability the system required.

The system also needed to work across the variety of conditions found in real care and rehabilitation environments: different floor surfaces, different footwear, patients with widely varying foot and motor-control disorders, and the mechanical variability of how different people walk. Calibration and adaptability were built-in requirements from the start.