Hook
Voyager 1 is 15 billion miles from Earth, still transmitting, still alive. But its power supply is dying—losing about four watts every year as the plutonium-238 in its generators decays. NASA can’t recharge it. They can’t retrieve it. They can’t send spare parts. All they can do is decide, one by one, which scientific instruments to shut down forever to keep the spacecraft’s core systems running a little longer.
This is the ultimate constrained resource problem. How do you manage something when replacement is impossible, repair is impossible, and every decision is permanent?
The Power Budget
Voyager runs on three radioisotope thermoelectric generators—RTGs. They convert heat from decaying plutonium-238 into electricity. In 1977, at launch, they produced about 470 watts combined. Today, they produce around 249 watts. By 2030, that number will drop to approximately 220 watts.
The math is unforgiving. Plutonium-238 has a half-life of 87.7 years, which means the generators lose roughly 0.78% of their output annually. That’s four watts. Every year, the power budget shrinks. And unlike a household budget where you can earn more or borrow, Voyager’s envelope only goes one direction.
This is what engineers call a power budget—the total energy available, allocated across all systems. The spacecraft needs power for: the radio transmitter (to send data back to Earth), the attitude control system (to keep the antenna pointed at us), the computer, the heaters (to prevent instruments from freezing), and the science instruments themselves.
When the budget shrinks below what all systems need, something has to go. And once you shut an instrument down in deep space, you can’t turn it back on. The cold destroys it.
Prioritization Framework
NASA uses a framework: science value versus power consumption, weighed against mission objectives that still matter.
In 2019, they shut down the cosmic ray subsystem—an instrument measuring high-energy particles from outside our solar system. It drew power, and its data, while valuable, was less critical than other instruments still measuring the space between stars. In 2023, they turned off one of the plasma science instruments. Earlier this year, they shut down another.
Each decision follows the same logic. First: Can we still achieve core mission goals without this? Voyager’s primary mission ended in 1989 after it passed Neptune. Everything since has been bonus—interstellar science. But not all interstellar science is equal. Measuring magnetic fields in the space between stars matters more now than measuring cosmic rays we’ve already characterized.
Second: Is there redundancy? Some instruments have overlapping functions. When that’s true, you keep the one with lower power draw or broader capability.
Third: What’s the power-to-science ratio? An instrument using three watts that returns unique data beats an instrument using five watts that duplicates what another sensor captures.
This is triage thinking. Hospitals do it in mass casualty events: immediate, delayed, expectant. Businesses do it in downturns: core operations, strategic bets, nice-to-haves. Infrastructure engineers do it with aging bridges: repair, monitor, replace. The framework is identical. Assess what you have. Rank by necessity. Cut what you must to preserve what you can’t lose.
Remote Impossibility
Every command sent to Voyager 1 takes 22 hours and 30 minutes to execute. The signal travels 11 hours and 15 minutes to reach the spacecraft. Voyager receives it, acts on it, and sends confirmation back—another 11 hours and 15 minutes.
You cannot try something and see what happens. You cannot course-correct quickly. You send a command, wait nearly a day, and hope your simulations were right.
This forces a specific kind of discipline. Before NASA sends any shutdown command, they run exhaustive simulations. They model the power budget with the instrument off. They check for cascading failures—does losing this sensor affect others? They test the command sequence in ground replicas of Voyager’s computer systems, which are identical twins built in the 1970s and maintained for exactly this purpose.
And they accept that once the command executes, it’s permanent. The instrument will freeze. Its calibration will drift. Even if they had surplus power years later, they couldn’t bring it back.
This is high-stakes, low-reversibility decision-making. Strategic pivots work this way—once you sell a division, you can’t easily reacquire it. Infrastructure choices work this way—once you decommission a rail line, rebuilding costs more than the original. Medical interventions work this way—some surgeries can’t be undone.
The principle is the same: when you can’t iterate, you simulate exhaustively and commit fully.
Graceful Degradation
Voyager was designed for this. The engineering term is graceful degradation—systems that lose functionality gradually rather than catastrophically.
Each science instrument on Voyager operates independently. Shutting one down doesn’t crash the others. The power system is modular—each RTG feeds into a shared bus, but the distribution system can isolate loads. The computer has redundant memory. The attitude control has multiple thrusters.
Graceful degradation means you plan for decline at the design stage. You don’t build monolithic systems where one failure kills everything. You build systems that can lose pieces and keep operating, even if diminished.
Web servers do this with load balancers—if one server fails, others handle the traffic. Buildings do this with firebreaks—if one section burns, the rest survives. Your body does this with kidneys—you can lose one and live.
The opposite is catastrophic failure. A single point of failure brings down the whole system. Early spacecraft had this problem—one sensor fails, the mission ends. Voyager’s designers learned from those losses. They built in redundancy, modularity, and the ability to shut things down without killing the core.
The result: a spacecraft designed in the 1970s is still operating in 2026, decades past its planned lifespan, because it can lose limbs and keep its heart beating.
When Replacement Is Impossible
Voyager 1 will never be retrieved. No future mission will catch it, repair it, or bring it home. Once its power runs out completely—likely sometime in the early 2030s—it will go silent. It will continue coasting through space, inert, for millions of years.
This changes the calculus. You’re not managing something you can fix later. You’re stewarding something irreplaceable through its final years.
This is true for more systems than we admit. The Arecibo Observatory in Puerto Rico collapsed in 2020—irreplaceable, built into a natural sinkhole, too expensive to rebuild. Legacy software running critical infrastructure—written in languages no one teaches anymore, with original developers retired or dead. The human body past a certain age—you can’t swap out organs indefinitely.
When replacement is impossible, the question shifts. Not ‘how do we fix this?’ but ‘how do we extract maximum value from what remains?’
For Voyager, that means prioritizing instruments that measure what we’ve never measured before. The magnetometer stays on because it’s mapping interstellar magnetic fields for the first time in human history. The plasma wave instrument stays on because it detects the boundary oscillations of the heliosphere—the bubble our sun creates in space.
For aging infrastructure, it means targeted maintenance on the highest-value sections, not blanket repairs everywhere. For failing business units, it means harvesting cash flow and customer relationships before shutdown, not pretending rejuvenation is possible. For end-of-life medical decisions, it means quality over duration, comfort over intervention.
The principle is the same: honest assessment of what’s left, ruthless prioritization of what matters most, and acceptance that some things end.
Close
Voyager’s engineers aren’t just managing a spacecraft—they’re demonstrating how to think clearly about anything finite, aging, and beyond repair.