Batteries are the tether that keeps most drones close to home. Every unmanned system that loiters, watches, or relays data eventually has to come back — or die — when its cells run dry. A DARPA-funded team led by Morgan State University wants to cut that tether entirely, by building a power cell that runs not on chemistry but on the steady, decades-long decay of nuclear waste.

The effort, called SYMPHONEE, is being developed under DARPA's "Rads to Watts" program, a $3.37 million contract awarded to a multi-organizational team that includes Northrop Grumman, Pacific Northwest National Laboratory (PNNL), ARA, Project Omega and Widetronix. According to a July 3, 2026 report from Defense One, the project targets a specific power greater than 10 watts per kilogram with a shelf life measured in years rather than hours, with a working prototype planned for early 2027 at PNNL.

What SYMPHONEE Actually Is

SYMPHONEE stands for Strontium-Yttrium Multi-junction PIN-based High-Density Output Nano-system for Extreme Environments, according to a June 15, 2026 Businesswire release announcing the contract award. The device is a radiovoltaic — a solid-state cell that converts the energy released by radioactive decay directly into electricity, without turbines, moving parts, or combustion.

The isotope at the center of the design is Strontium-90, a byproduct of nuclear fission. Rather than treating it purely as waste to be stored, the SYMPHONEE approach layers the isotope directly against a semiconductor structure so that the radiation it emits is converted into usable current, according to a technical write-up from Interesting Engineering. That publication describes the design as solid-state, with no moving parts and no refueling requirement — a profile that, if it holds up, would make the cell fundamentally different from both batteries and mechanical generators.

Who's Building What

The Businesswire release lays out a clear division of labor across the team. Morgan State University serves as the prime contractor — notably, the release states this makes Morgan State the first Historically Black College or University (HBCU) to lead a DARPA project. PNNL is responsible for handling the nuclear materials and conducting testing. Project Omega is building the actual nuclear power generator. Widetronix is designing the semiconductor power converter that turns the decay energy into usable electricity. Northrop Grumman and ARA contribute computational modeling to the broader effort, per Defense One's reporting.

The DARPA Program Behind It

SYMPHONEE sits inside DARPA's "Rads to Watts" program, described on the agency's official program page as an effort to build radiovoltaics that convert nuclear radiation directly into electricity for unattended power in extreme, radiation-heavy environments. The program is structured as an Other Transaction for Prototype (OT-P) agreement with a 15-month base period, and DARPA's stated performance targets are power density, measured in watts per square centimeter, and specific power, measured in watts per kilogram — the same specific-power metric SYMPHONEE is chasing at its greater-than-10-W/kg goal.

Q&A: The Basics

Is this a nuclear reactor on a drone?
No. The sources describe a radiovoltaic cell, not a reactor — there's no fission chain reaction taking place aboard the device. It relies on the passive decay of Strontium-90, converted directly into electrical current through a semiconductor structure, per Interesting Engineering's description of the design.

How long would it last?
Defense One's report describes a "yearslong" shelf life, and Interesting Engineering frames the program's aim as a power source capable of delivering electricity for multiple decades without refueling. DARPA's own program page frames the broader goal as unattended power for extreme environments where resupply isn't practical.

When will there be a working prototype?
Defense One reports the team is targeting a working prototype in early 2027, to be built and tested at PNNL. The next 18 months of the program are focused on reducing technical risk before that milestone.

What would it power?
According to Defense One's and Interesting Engineering's reporting, the applications under consideration span drones, satellites and other space systems, underwater infrastructure, and remote sensors — categories of systems that share a common problem: they often operate in places where swapping batteries or refueling generators is difficult, dangerous, or impossible.

Why It Matters

Endurance is the single biggest constraint on what an unmanned system can do. A drone that can fly for hours instead of minutes, or a remote sensor node that can run for years instead of weeks, doesn't just do the same job longer — it changes what missions are even possible. Persistent surveillance, long-dwell maritime patrol, and sensors left in denied or hard-to-reach terrain all live or die on power budgets.

A radiovoltaic cell with no moving parts and a multi-decade lifespan, if SYMPHONEE hits its targets, would sidestep the two things that currently limit unattended systems: battery degradation and the logistics of resupply. That's precisely the gap DARPA's Rads to Watts program is aimed at — unattended power in environments where nobody is coming back to change the batteries. For the drone industry specifically, the interest from Northrop Grumman and the explicit mention of drone applications in Defense One's reporting signal that this isn't purely a lab curiosity; it's being evaluated with airborne platforms in mind from the outset.

It's also worth noting what SYMPHONEE is not, at least based on what's been disclosed so far: it is not a chain-reacting reactor, and the publicly available reporting doesn't specify a total energy output or an explicit power-vs-flight-time tradeoff for any particular airframe. The team's own stated near-term goal is modest and specific — hit a power-density target and reduce technical risk over an 18-month window — rather than a finished, flight-ready power system. Whether a sub-watt-scale radiovoltaic cell can be scaled into something that meaningfully extends a drone's flight time, or whether it's better suited to trickle-charging low-power sensors and avionics over years, is exactly the kind of question the 2027 prototype is meant to start answering.

There's a symbolic dimension too: Morgan State's role as prime contractor marks, per the program announcement, the first time an HBCU has led a DARPA project — a notable data point in a defense-research funding landscape that has historically concentrated prime contracts among a small set of institutions.

For now, SYMPHONEE remains a proof-of-concept under a 15-month base agreement. The real test comes at PNNL in early 2027, when the team's design has to move from watts-per-kilogram targets on paper to an actual cell converting Strontium-90 decay into usable current.

Sources