How NASA saved Voyager 1

Our beloved spacecraft Voyager 1 was dead. It would never again send back usable science data.

At least, that’s what the Voyager 1 team feared — but with some ingenious interstellar troubleshooting and firmware updates from 15 billion miles away, NASA’s team at JPL pulled off a miracle.

Here is how NASA saved the Voyager 1 spacecraft.

The problem: Voyager 1 was speaking in gibberish

On November 14, 2023, Voyager 1 started acting weirdly. The spacecraft, which launched in 1977, communicates with Earth in binary code — a series of 1s and 0s. This is normal. But this binary was indecipherable — it was basically gibberish, a repetitive sequence of ones and zeroes with no meaning. It looked like the computer was stuck.

The Voyager 1 team out of NASA’s Jet Propulsion Laboratory, or JPL, in Pasadena, California, determined that the problem was with the FDS, or flight data system. The FDS consists of a CPU, memory chips to store code and data, and interfaces to receive and transmit data from other systems on the spacecraft.

One of the FDS’s main jobs is to receive raw data from Voyager 1’s science instruments as well as engineering data on the health of the spacecraft. It puts this all into a neat package, and then sends it over to the TMU. or telemetry modulation unit, for transmission back to Earth. With the TMU sending back gibberish, scientists were able to pinpoint that the issue was likely in the FDS sending bad data to the TMU.

Because Voyager 1 is in interstellar space, it takes 22.5 hours each way — that’s 45 hours roundtrip — to communicate with the spacecraft. It’s part of why troubleshooting is such a challenge.

In December, the Voyager 1 team sent a signal for the spacecraft to restart the FDS, in hopes it would return that computer to a previous state, before the problems began. It didn’t work.

And keep in mind, besides the distance issue, the team who’s currently diagnosing and troubleshooting the spacecraft is relying on 50-year-old documentation written by people who had no clue that the spacecraft would be operating this far in the future, nor what computers would look like. They have to test and re-test commands before sending them because the team doesn’t really know what unintended consequences might result from sending commands to Voyager 1. They have to be sure they know exactly what they’re doing. And things weren’t looking good.

The Voyager 1 team started to communicate with the public that this might be the end of the spacecraft. It would continue through interstellar space, but it was looking increasingly possible we’d never receive meaningful data from Voyager 1 again.

They continued troubleshooting for months, working on the problem and trying to figure out what was wrong with our beloved spacecraft. The spacecraft was clearly receiving commands, and it was sending a steady stream of data back to Earth. The problem was just that none of it was usable data.

But in March, there was a breakthrough: On March 1, the Voyager team sent what’s called a “poke” to the spacecraft. The working theory at that point was that some of the FDS’s memory was corrupted, and this poke was designed to instruct the spacecraft to try and work around any corrupted sections of its memory by attempting different sequences within its software. Well, the team got the same gibberish signal back on March 3 — but alongside it was another signal. On March 7, they realized this second signal wasn’t gibberish at all. And on March 10? They realized this second signal contained an entire readout of the FDS memory. This was the first positive sign we’d gotten that maybe we could save Voyager 1.

A quick history of the twin Voyager spacecraft

This wasn’t the first time there were problems with these two — we’ve gone months without communication with these two. Back in 2020, Voyager 2 dropped out of contact for months because upgrades to the Deep Space Network took the only dish at the time capable of talking to that spacecraft offline. Back in 1981, there was a similar computer problem to the one I’m telling you about (though it’s unrelated to this issue).

Voyager 1 and 2 are almost 50 years old. That’s ancient for a spacecraft. You know, when I talk about the Hubble Space Telescope’s problems (it’s currently in one gyro mode, which is the last phase of usable life for this beloved observatory) because it’s aging, it launched in 1990. That makes it around 34 years old. In contrast. Voyager 1 and 2 launched in 1977.

The mission was originally to explore the outer planets. The twin spacecraft were built for a five-year mission, though there was always hope they would endure beyond that original timeline.

The original vision was to take advantage of a rare planetary alignment of Jupiter, Saturn, Uranus, and Neptune, which occurs every 175 years, and get a four-planet tour for a minimum amount of propellant. However, building a spacecraft for those specifications, and for that primary mission, was deemed to be too expensive, so Voyager 1 and 2 were built and funded for just a two-planet mission, to study Jupiter, Saturn, and their moons.

Voyager 1’s trajectory took it through the asteroid belt and then it flew by Jupiter on March 5, 1979. It also approached the Jovian moons Io, Europa, Ganymede, and Callisto.

After this flyby, the JPL team adjusted Voyager 1’s trajectory so it would fly by Saturn and its moons Titan, Tethys, Mimas, Enceladus, Rhea, and Hyperion. Voyager 2’s trajectory was designed to preserve the Uranus and Neptune flyby option, and JPL was able to take advantage of that because of how well the spacecraft was doing.

But for Voyager 1, that was the end of its flybys. The gravitational assist from Saturn sent the spacecraft on a course outside the solar system. Voyager 1 entered interstellar space on August 25, 2012. Interstellar space is basically the space outside our heliosphere — outside the influence of the sun and its solar wind of charged particles that surrounds the sun and planets. Voyager 1 was the first humanmade object to enter interstellar space.

The spacecraft is currently located in the constellation of Ophiuchus, around 15.1 billion miles, 24.3 billion km, or 162.7 AU( or astronomical units) away from Earth. It’s traveling at around 38,000 mph or 61,000 kph, and it’s currently between the Kuiper Belt (which extends to the outer edge of our solar system) and the Oort cloud (which it won’t enter for another 300 years or so). For more on the Kuiper Belt and the Oort cloud, check out my newsletter.

The daring fix for Voyager 1

It’s highly unlikely Voyager 1 will be in communication with Earth long enough for us to receive science data about the Oort cloud. But could we get just a few more years out of it? That’s what scientists were working on.

In late March, scientists finally determined that the problem with Voyager 1 was indeed a corrupted section of FDS memory. The readout they received on March 3 contained both the software code as well as variables that indicate the spacecraft’s status. Through this FDS memory readout, they were able to determine that around 3 percent of the FDS memory had been corrupted. The working theory at this point was that a chip had gone bad — a hardware issue, rather than a software one. Whether it was just due to age or a bad hit from an energetic particle wasn’t clear, but they had to figure out a way to fix it.

It took awhile, because remember what I said before, every single command sent to the spacecraft could have unintended consequences, and they have to be VERY careful. In 2023, a command sent to Voyager 2 inadvertently caused it to point its antenna slightly away from Earth and we lost contact with it for a few weeks. They have to be very precise.

The team spent their time figuring out a way around the corrupted memory — basically crafting an intricate set of code instructing Voyager 1 where to store its engineering data.

This is more complicated than it seems: Remember this is a computer from the 1970s. Voyager 1 has less than 70 kb of memory in its entire computer — not just the FDS, the entire computer. Finding places to store new data when 3% of the memory available is corrupted is a challenge.

And they weren’t actually able to find one place to store all the spacecraft’s engineering data. Instead, they had to find available space and then program code that would divide the engineering data between these spaces, but ensure that the code was written such that each piece new where to find the next piece so it would all function seamlessly as one unit. All from 15 billion miles away with 22.5 hours of communication travel time each way.

The team sent the new instructions to Voyager 1 on April 18. On April 20, they received the first meaningful communication from Voyager 1 since November, with all of the engineering data intact.

Getting the science instruments back online

But this isn’t the end of the story. They had all the engineering data, and were able to look at the health of the spacecraft — but Voyager 1 still has four working science instruments.

When Voyager 1 launched in 1977, it had 10 total instruments. There are currently four that are operational.

That’s because the spacecraft’s power is also dwindling. Voyager 1 uses three Radioisotope Thermoelectric Generators (RTGs) for power. The U.S. has been using them since 1961, and they’re a great option when solar power isn’t really feasible, as it wasn’t on Voyager 1 due to the distance it would travel from the sun. RTGs rely on the radioactive decay of the plutonium-238 isotope for electricity (no, this is not the kind of plutonium used in nuclear weapons). But the problem is that as the plutonium decays, there is less and less power available for the spacecraft. The Voyager 1 team has had to strategically switch off instruments and heaters in order to conserve power as long as possible.

The current estimates say that the longest Voyager 1 will be able to send back science data is 2030, and we will likely continue to receive engineering data through 2035. After that point, the spacecraft will continue its journey through the cosmos, but it will no longer be able to communicate with Earth.

The four operational instruments aboard Voyager 1 are the cosmic ray system, the magnetometer, the low energy charged particle instrument, and the plasma wave subsystem — and the Voyager 1 engineering team had to repeat the engineering data code process with each of these science instruments in order to get usable data back.

On May 17, the science team sent the first of those instructions. Two days later, they received the first science data in six months from the plasma wave subsystem and magnetometer. Two down, two to go.

And on June 13, we got the final piece of good news: It apparently took some additional work, but all four operational science instruments for Voyager 1 are now working correctly and sending usable data to Earth.

There’s still some maintenance to do, which isn’t a surprise: It’s a 46-year-old spacecraft. But it’s good to be at a point where all this spacecraft needs right now is maintenance.

We’re going to have to say goodbye to Voyager 1 eventually. That time is near. But it’s nice that we haven’t had to say goodbye yet, and it’s good to see the team cares so much about saving this little spacecraft that they pulled off the near impossible.

skrishna

View more on Instagram