Humans are wired for short-term gratification and comfort-seeking behaviors. As documented by a Harvard University dissertation study of The Marshmallow Test, there are no significant cognitive, behavioral, and demographic factors that correlate with delayed gratification. The significance of pleasure-seeking behavior is relevant to wearable device designers because sensor accuracy and clinical validation don’t matter if patients don’t pick up, put on, or otherwise proactively engage with the devices. You can’t push an alert button after a fall if you aren’t wearing the device and you can’t track your steps if you don’t put on a fitness wearable. Advances in passive sensor technologies, however, may be the best answer to patient noncompliance.

Emerald Innovation, an artificial intelligence analytics startup by students from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) is developing non-contact wireless sensor technology that works with machine learning algorithms to overcome patient adherence issues. Emerald published a paper that demonstrates that its through-wall human pose estimation is “almost as accurate as the vision-based system used to train it.” Emerald’s tech employs radio signals to estimate 2D poses. Frequencies in the Wi-Fi range bounce off human bodies, allowing human movement monitoring. Emerald uses RF-Pose, a neural network with a wireless signal 1,000 times less powerful than Wi-Fi but still capable of estimating human skeleton position through walls. According to Emerald, the AI data derived from RF-Pose can reveal mobility, gait, daily living activities, respiration, sleep quality, and sleep stages.

Emerald is currently developing applications that use its passive through-wall sensing and AI metrics with sleep disorders, musculoskeletal disorders, pulmonary disease, and neurological diseases. Applications of similar technologies were developed at MIT in the early 2000s for intelligence and military applications, but as wireless sensing and AI continue to improve, passive sensing may be the best response to patient noncompliance.