DARPA's 10W/kg batteries
DARPA is pushing compact nuclear‑power concepts that claim about 10 watts per kilogram — a density that could radically extend endurance for drones, sensors and battlefield electronics. (x.com) If those power figures scale safely, it would change design tradeoffs in engineering smaller long‑endurance platforms that today rely on bulky fuel or limited batteries. (x.com)
A normal battery stores chemical energy the way a pantry stores food: it starts full, and every watt you pull out empties it a little more. A radioisotope power source works differently, because the fuel keeps releasing energy on its own for years as atoms decay. (science.nasa.gov) The version most people have already seen is the radioisotope thermoelectric generator on NASA rovers like Curiosity and Perseverance. That machine turns heat from plutonium-238 decay into electricity and makes about 110 watts at the start of a mission. (science.nasa.gov) DARPA is chasing a different design. Its Rads to Watts program is focused on radiovoltaics, which try to turn nuclear radiation directly into electricity the way a solar cell turns light directly into electricity. (darpa.mil) That direct-conversion idea sounds simple, but the hard part is damage. The same radiation you want to harvest is also smashing the semiconductor material, and DARPA says those radiation-induced defects are what degrade the performance and lifespan of current radiovoltaics. (darpa.mil) DARPA’s target is not a tiny coin-cell trickle charger. The program’s stated end goal is radiovoltaics that can convert high-power nuclear radiation into kilowatts of electrical energy. (darpa.mil) The news this week is that Seattle startup Avalanche Energy said on April 8, 2026 that it won a $5.2 million, 30-month DARPA contract under that program. Avalanche says it will build solid-state, micro-fabricated cells that convert alpha particles from radioisotopes directly into electricity. (avalanchefusion.com) Alpha particles are heavy charged particles, basically helium nuclei flying out of radioactive fuel. Because they carry charge, a device can try to catch their energy directly instead of first turning it into heat and then running that heat through a generator. (osti.gov) Avalanche says the system is being designed for more than 10 watts per kilogram. That is a specific-power figure, which is the engineering version of asking how much power you get for each kilogram you have to carry. (avalanchefusion.com) That number stands out because NASA’s current Multi-Mission Radioisotope Thermoelectric Generator produces about 110 watts at launch from a unit that weighs about 45 kilograms, or roughly 2.4 watts per kilogram. DARPA’s threshold is therefore about four times that specific power. (mars.nasa.gov, science.nasa.gov) DARPA is also pushing performers on the ugly engineering details, not just the headline number. In its published program answers, the agency says teams can count creative waste-heat management toward the 10 watts-per-kilogram estimate, but they will still be downselected on the measured performance of the radiovoltaic cell itself. (darpa.mil) If a few-kilogram package can really deliver laptop-class continuous power for months, the design math changes for small spacecraft, unattended sensors, and some drones that today have to choose between short battery life and much bulkier fuel systems. Avalanche says it will test resilience with both particle accelerators and active radioisotopes, which is the part that decides whether this stays a lab demo or becomes hardware people trust. (avalanchefusion.com, darpa.mil)
