Four disciplines, practiced together

Devices that people keep on. We design multisensor wearables that noninvasively collect a suite of biosignals and biomarkers — the sensing, the packaging, and the human factors that decide whether a device is worn for an hour or for a year. This is the work behind a granted patent, a pending patent, and a company.

• Multisensor biosignal and biomarker acquisition
• Continuous, distributed monitoring architectures
• Form factor, fit, and wearability
• Power budgeting for continuous wear

Machines that move with a person rather than against them. Our robotics work spans wearable robotic devices for rehabilitation, prosthetics, and exoskeletons, including the upstream biomechanics needed to make them behave in a human-like way.

• Wearable robotics and exoskeletons
• Prosthetic device design
• Actuation and mechanism design
• Biomechanical characterization and modelling

The layer where most early-stage devices quietly fail. We build the embedded systems and sensor integration that keep a device accurate once it leaves controlled conditions — and we specify them alongside the mechanics rather than after them.

• Firmware and embedded systems
• Low-power design
• Sensor selection and integration
• Data acquisition pipelines

Raw biosignals are rarely the thing you want to measure. We build the signal processing and AI models that turn correlated, noisy, drifting inputs into an estimate worth acting on — and we are candid about when a model is covering for a sensing problem that should be fixed in hardware.

• Biosignal and biomarker extraction
• AI models for physiological estimation
• Continuous monitoring and inference
• Validation against ground truth