Is there any reason that NASA's launched space probes and MERs usually operate decades longer than expected?

Question

Space Probes & MERs History Recap

---

Pioneer 10 & its twin probe Pioneer 11 ^reference

Courtesy of Uwe's response on my question :-

• Pioneer 10 was planned for 21 months of operation. It did work 31 years, that is 16 times longer. We should be very happy about such a reliable space probe.

Voyager 2 & its twin probe Voyager 1 ^reference

• Voyager 1 and 2 were designed for 5 years of operation: They work 41 years now, about 8 times longer. So Pioneer 10 did work 16 times longer than planned, much longer than the Voyagers. If the Voyagers work till 2025, it will be only about 10 times longer than planned.

MER-A ^{known as Spirit}

• The rover completed its planned 90-sol mission. Aided by cleaning events that resulted in more energy from its solar panels, Spirit went on to function effectively over twenty times longer than NASA planners expected. Spirit also logged 7.73 km (4.8 mi) of driving instead of the planned 600 m (0.4 mi), allowing more extensive geological analysis of Martian rocks and planetary surface features.

MER-B ^{known as Opportunity}

• With a planned 90 sol duration of activity (slightly more than 90 earth days), Spirit functioned until getting stuck in 2009 and ceased communications in 2010, while Opportunity was active as of June 10, 2018 when a dust storm forced it to hibernation. It has operated 5275 sols since landing, having exceeded its operating plan by 14 years, 215 days (in Earth time). Opportunity has operated for over 55 times its designed lifespan. As of June 10, 2018 when contact was lost, the rover had traveled a distance of 45.16 kilometers (28.06 miles).

Nonetheless, with so many references showed that the Pioneer program, Voyager program and two relevant MERs Spirit & Opportunity were so successful and already operated for over dozen times longer than initially expected. It is undeniable that NASA is indeed a great and professional national space agency. But presumably those space probes & MERs were actually designed to operate that long in the first place, why wouldn't NASA just announces their expected lifespan slightly off than their actual lifespan instead; those expected and actual numbers recorded in the history were so distant? Or should I say it is essentially unpredictable?

Maybe we can assume NASA didn't want to take unnecessary risks plus considered other external factors that would concerns the mission (e.g. decay of the RTG over time be more of an absolute limiting factor) so in a nutshell they'd rather purposely give a smaller number on expected lifespan. What's up with all these?

Answer 1

The idea that NASA would intentionally under-represent the lifetimes of those (cherry-picked) missions implies that somehow they knew those lifetimes before launch. Nothing could be farther from the truth.

You can't just calculate a lifetime. There is extensive testing and associated analyses performed on the hardware to provide sufficient confidence that it will last the required lifetime, i.e. to complete the primary mission. That testing is expensive, so no more is conducted beyond exactly what is needed to gain that confidence. Since they don't do testing beyond that, they have no idea how much longer the hardware might last.

A simple example is the wheel motors on MER. They are brushed motors, and would be expected to fail with some amount of use. The design was tested to three times the expected usage, with temperature swings, in order to have sufficient confidence in the motor lifetime. Typically there are failures during such testing, and design changes at made to address those failures, followed by retesting.

As I recall, the requirement for each rover was to go 600 meters. Opportunity has gone over 45 km. However one of the motors (a steering motor) did in fact fail many years ago. And a few years into Spirit's mission, one of the drive motors failed, leaving one wheel dragging a trench as it drove. That failure was a contributor to Spirit's demise, since it got stuck and could not move to a better angle for the Winter Sun. So the motors have lasted much longer than they were tested for. But no one knew that they could last that long, since they were never tested for it. And there were indeed failures of the motors, just not very many.

On the Voyagers, the scan platform, also dependent on mechanical devices, failed after the primary mission, requiring slewing the entire spacecraft for the extended mission encounters of Uranus and Neptune by Voyager 2. Also in 1978 Voyager 2 had one radio receiver completely fail, and the backup receiver had a capacitor fail preventing the receiver from tracking the frequency of the signal from Earth. Then the operations team had to develop special procedures to transmit to Voyager in exactly the frequency it could receive without tracking, taking into account the Doppler shift. Those receiver failures occurred during the primary mission.

There are many others like that. Much of what you see in the success of these missions are due in part to workarounds of various failures that occur in both the primary and extended missions.

No one knows how long these things will live before they live that long. And often the reason they can keep going is ingenuity on the part of the operators.

I can assure you that no one is more surprised at the longevity of these missions than the people who built them.

Answer 2

Taxpayers are really unhappy when expensive missions fail before achieving the promised goals. Due to politics, inside NASA, failure is not an option.

Every device has a MTBF estimate. Failure prediction for a system is an engineering discipline and problem. To calculate the total MTBF for a spacecraft, each tiny part's MTBF is added up (slightly more math than addition needed), so estimates for each part, component, assembly, & system are included. In the end, a MTBF is calculated for everything in the mission payload. Due to complexity of these devices, most of the parts will need to be slightly over engineered for the total system to work for at least the required timeframe or cycles. This over-engineering is done because you cannot swap out a $5 part from 20M miles away or almost anywhere in space.

In practice, the result is that space craft tend to work beyond their minimum lifespans and if there is a failure in 1 area, often another area can handle the extra duty to continue with the mission.

You might remember when the Hubble Space Telescope was first launched and found to have optics problems which weren't discovered until on-orbit and how foolish all of NASA looked to taxpayers, Congress and the world. Because the entire assembly wasn't put together on the ground and validated to save money and the timeline (also due to political pressure), multiple shuttle missions were required to grab success from the perceived failures. They also upgraded some of the components which make HST better than originally designed.

Only a manned mission could have performed all the required upgrade tasks.

As long as we send 1 probe out per mission, it will necessary be over engineered, slightly. If we switch to sending out 10-100 smaller probes, each of them will provide a level of redundancy so the over engineering will be reduced and we will see some failures and some continue beyond their expected lifespans. The same has happened with home electronics and other devices. Cheaper device usually don't last as long as the more expensive products.