Biomedical Technology · us
Helping Patches Read the Body’s Signals: ARPA-H Commits Up to $117.4 Million to Develop Modular Biosensors
From heart failure and inflammation to the menopausal transition, four teams will explore wearable patches that continuously measure biomarkers. The real test is whether the signals collected over time can be turned into reliable information usable in care.
A single test can capture the body’s state at a particular moment, while continuous measurement offers a chance to see the direction of change. For patients with heart failure, people recovering at home after a myocardial infarction, or women undergoing the menopausal transition, these changes could offer new clues for care. The U.S. Advanced Research Projects Agency for Health (ARPA-H) announced on October 1 that its Delphi program will commit up to $117.4 million over four and a half years to support four teams developing next-generation wearable biosensors; funding disbursements are conditional on meeting research milestones.
The program centers on dividing sensors into recombinable “chiplets”: specialized components handle biomolecule recognition, power management, signal processing, and data transmission, and are then assembled for different uses. If the components can interoperate, adding new measurement capabilities in the future could allow teams to reuse an existing platform, reducing the burden of designing an entire device from scratch each time.
The Massachusetts Institute of Technology’s PYTHIA project, led by Ahmad Bahai, has been awarded up to $37.9 million and aims to continuously monitor biomarkers related to heart failure using a wearable patch. The team plans to power the device with heat generated by the patient’s body and incorporate low-power artificial intelligence on the chip. AI here is part of the sensing device’s design; publicly available information has not yet explained how the model will be trained or what data it will use, nor has it provided accuracy figures for detecting changes in the patient’s condition.
New York University is moving the measurement site beneath the skin. The SENTINEL project, led by Elisa Riedo, has been awarded up to $30.8 million and will combine miniature probes with DNA sensing elements to continuously track inflammatory signals, while attempting to harvest energy from radio-frequency sources such as Bluetooth and Wi-Fi. The design addresses both molecular recognition and power supply, but whether it can maintain stable measurements in everyday use remains to be confirmed through prototype testing.
A University of Washington team led by Quansan Yang has been awarded up to $7.5 million and has chosen to track the menopausal transition through sweat. The patch will use engineered proteins to recognize biomarkers and adjust the local pH around the sensor to restore the proteins to a reusable state, extending its sensing capability. The public description has not yet listed the specific molecules to be measured or provided validation results linking sweat signals to clinical states.
Novelna’s PULSE project, led by Ashkan Afshin, focuses on monitoring inflammation after myocardial infarction, with the hope of helping reduce patients’ chances of rehospitalization. The team plans to use the human body as a channel for data transmission to reduce transmission energy consumption and extend battery life. This project and MIT’s heart failure patch address different care settings, and reducing rehospitalization rates remains a development goal rather than a proven benefit.
All four awards have a start date of September 24, 2026. Under the Delphi timeline, the teams must demonstrate full interoperability among components within the first 18 months and meet post-quantum secure communication standards; by month 24, they must demonstrate the first in vivo prototype. Before month 54, minimally invasive wearable devices must complete clinical trials, while noninvasive devices must complete human factors studies evaluating how users interact with the devices.
This announcement sets out research and development funding and a validation roadmap, without accompanying data on measurement accuracy in humans or clinical benefits. Once biological signals are collected continuously, how fluctuations are interpreted, when intervention is needed, and whether long-term wear is safe and comfortable will all affect whether these patches can enter care. Human factors studies can answer questions about use, but cannot by themselves establish that disease monitoring is effective; each device must still build evidence appropriate to its intended use to turn more data into better medical judgments.