A satellite drifts through orbit with no one able to plug it in. Its mission ends the moment its battery does. The Defense Advanced Research Projects Agency is funding a different idea: a solid-state cell that turns the steady decay of a radioactive isotope into electricity, the same way a solar panel turns sunlight into power, and keeps doing it for decades.
The program is called Rads to Watts, and its working target is a power cell that delivers more than 10 watts per kilogram, sits on a shelf for years, and runs for up to 30 years once deployed — a battery replacement rather than a chemical battery, as DARPA's program page puts it. A minimally viable prototype is expected by early 2027.
The mechanism is not new in physics. Radioisotope power has already kept deep-space probes alive for decades and runs the ionization chamber in a household smoke detector. What is different is the form factor. Traditional radioisotope generators, called RTGs, harvest heat from a decaying isotope and run it through a thermoelectric or Stirling converter. Rads to Watts instead aims to convert the radiation itself into a current inside a solid-state device, closer to a betavoltaic cell than to a thermoelectric brick, as the DefenseOne write-up describes. That is why a 30-year lifetime becomes plausible. There are no moving parts, no chemistry to break down, just a sliver of isotope pushing charge across a junction for as long as its half-life allows.
What the technology would actually buy is mission time. A drone that stays aloft for months without landing to swap a pack. A remote sensor on a polar ice shelf or a deep-sea cable that does not need a service boat. A satellite that outlives the orbital mechanics that doom its chemical cousins. Stafford Sheehan, the founder of participating vendor Project Omega, frames the device in those terms, telling DefenseOne that a small cell could replace traditional batteries by acting as a mini-generator sized for platforms that cannot be recharged in the field.
The money is already moving. Morgan State University holds a $3.37 million contract as prime on at least one workstream, with a proof-of-concept device funded under that award, per the SBIR.gov portfolio entry. Project Omega is positioning itself as a radioisotope mini-generator startup, and Chemical & Engineering News profiled the company in February 2026 as a stealth entry into small nuclear power. Established betavoltaic vendors such as Widetronix already sell micro-watt and milli-watt radioisotope cells, suggesting there is a small industrial base for the technology Rads to Watts is trying to scale up.
Adjacent teams are also working the same problem from a different angle. Interesting Engineering reports on a separate DARPA-funded effort at Symphonee that uses spent-fuel isotopes in a similar direct-conversion architecture, a useful reminder that Rads to Watts is not the only path being tested. Peer-reviewed underpinning for solid-state radioisotope conversion exists too, in the form of mechanism work published in Nature Scientific Reports.
None of this is a product yet. The 10-watts-per-kilogram target is a DARPA figure of merit, not a measured number on a finished cell. The 30-year lifetime is a program aspiration, not a demonstrated run time. A working prototype in early 2027 will still be years from a flight-qualified part. The physics says it is possible; the engineering has not yet shown it can be built at drone scale, packaged for storage, and qualified for the environments where it would actually fly.
The watch item, then, is the early-2027 prototype. If a Rads to Watts cell hits the 10 W/kg line and survives a multi-year shelf test, the conversation shifts from interesting physics to mission redesign around a power source that outlasts the platform it sits on. If it does not, Rads to Watts joins the long list of long-duration power ideas that the physics said was possible and the engineering never quite delivered.