A patch powered by body heat, microscopic probes beneath the skin, a sweat sensor aimed at perimenopause and a monitor for inflammation after a heart attack are now competing for a place in the next generation of wearables. ARPA-H announced the four teams on October 1, under a commitment of up to $117.4 million over 4.5 years. The announcement is not clinical proof, and no consumer launch has been promised. ARPA-H kicks off work to advance next-generation personalized biosensors
Four bets, not one miracle device
MIT is proposing a clinical-grade heart-failure patch powered by body heat. NYU is working on microscopic probes under the skin with DNA-based sensors and energy harvested from radio-frequency sources such as Bluetooth and Wi-Fi. The University of Washington is pursuing a non-invasive sweat sensor for perimenopause biomarkers, while Novelna is targeting inflammation after myocardial infarction and communication through the body itself. ARPA-H kicks off work to advance next-generation personalized biosensors
The shared idea is modularity: separate chiplets could handle biological recognition, signal processing, power, communications and security, then connect for different medical uses. ARPA-H says the teams must demonstrate interoperability and adherence to post-quantum secure communication standards by month 18. Reusable modules could reduce the need to rebuild an entire wearable for every biomarker, but that ecosystem remains a program goal, not a demonstrated platform. ARPA-H kicks off work to advance next-generation personalized biosensors Delphi: Developing self-monitoring ecosystems to 'know thyself'
The body is the difficult test bench
A sensor can detect a chemical and still fail as a medical instrument. Sweat does not automatically mirror blood, tissue conditions change, and readings can drift as a person moves or material builds up on the sensor. A September 2026 study described calibration drift and physiological variability as central problems and found only moderate correlations in the sweat-serum comparisons it examined. A technical review also identifies biofouling, sample transport, unstable biological receptors, regeneration and on-body calibration as major barriers. Physics-informed machine learning for robust calibration and physiological validation of wearable electrochemical sweat biosensors for metabolite monitoring Wearable biosensors for healthcare monitoring
That gap has direct consequences. The FDA has warned that unauthorized watches and rings claiming to measure blood glucose can produce inaccurate readings and lead to dangerous treatment decisions. The warning concerns those products, not Delphi’s proposed devices, but it shows why a convenient form factor cannot substitute for analytical and clinical validation. Do Not Use Smartwatches or Smart Rings to Measure Blood Glucose Levels: FDA Safety Communication
The deadlines will decide what survives
ARPA-H calls for a first in-vivo prototype by month 24. By month 54, teams are expected to reach either a clinical trial for minimally invasive wearables or a human-factors study for non-invasive devices. Those milestones may narrow, delay or end individual projects; they do not guarantee approval, reliable diagnosis or treatment benefit. ARPA-H kicks off work to advance next-generation personalized biosensors
The program’s wager is not that a watch will soon read every hidden change in the body. It is that interchangeable hardware can make difficult sensors easier to build and test. The decisive evidence will arrive when these proposed modules meet living tissue, repeated measurements and clinical decisions.
