Leveraging machine learning for arrays of soft sensors

Sensor arrays are ubiquitous. They capture images in digital cameras, record the swipes of our fingers on the screens of our phones and tablets, or map pressure distribution over an area. Soft capacitive sensors have long been proposed to make electronic pressure-sensing skins. However, although different designs of entirely soft capacitive sensors have been proposed, large arrays of those sensors are challenging to produce. Indeed, arrays require high-resolution patterning of electrodes, and routing of long and thin electrical connections. These two tasks remain difficult or costly for the high-resistivity compliant electrodes of dielectric elastomer sensors. Instead of relying on the complex patterning of arrays to provide location resolution, we propose to use a plain, unstructured sensor with a single pair of electrodes but rely on computing power to infer pressure location and amplitude from clever sensing signals. Here, we propose a new machine-learning-based approach, which enables us to identify pressure location on a continuous 1D sensor split into 5 sensing zones with an accuracy greater than 98 %. We also demonstrate that we can identify pressure location and qualitative pressure magnitude (soft, medium, hard) on a 3-zone sensor with 99% accuracy.