NASA’s got a history of their rovers overperforming like this. Between this and Webb, it’s hard to argue they’re not the best out there.
(*I don't necessarily mean the practice of shooting a rocket into space and launching a telescope, but the ability to build an object that can perform that exactly to its specifications under those kinds of circumstances.)
Nasa, congrats on successful mission, which now ended.
I see what you did there.
IMO, Webb is a bad example. It was super over budget, late, and experienced a litany of problems.
In contrast, most un-manned NASA projects are roughly on budget and generally have few issues. And even the ones that do, NASA often figures out how to correct them like the Hubble telescope.
This is simply not true. Maybe for some of the smallest Discovery or New Frontiers missions, but then those aren’t actually built by NASA/JPL.
Mars Sample Return is also famously insanely late and over budget, and so have been most flagship missions.
JWST was built by primarily Northrop Grumman engineers
Give other space agencies the budget of NASA, let experience marinate and you will see similar performance. NASA does not have some sort of magical intrinsic property that makes engineers perform better by sheer effect of working under its umbrella.
For example, what requirements for the Ingenuity helicopter were set for the engineers? Five flights or a 100 of flights?
Unfortunately this leads to uninformed reporters to make stupid claims like "a probe only designed to last 90 days...". There's no self destruct or planned obsolescence in NASA probes. They don't have an egg timer set to 90 days and then explode.
The probes are designed to last at least as long as the ability for their consumables/power supply lasts. Consider the stresses of launch, transit through space, and insertion/landing are way worse than day to day conditions for a lot of probes. If they can survive all of that rolling around a desert is relatively easy (in comparison). If Ingenuity made one successful flight there was little reason it couldn't make more so long as it had the appropriate power available and the Martian Space Defense didn't shoot it down.
Ingenuity didn't overperform, it performed according to its actual design and construction. It only overperformed misinformed media expectations.
C'mon let's not blame the reporters. NASA pushes this line - why do think everyone repeats it, using the same language almost every time? You can even see it here on HN.
NASA never wants to say a probe will probably last longer than its primary mission because then reporters go the complete opposite direction and claim NASA said a "probe failed to last the claimed X+Y flights".
Under promise, overdeliver, is a nice way to say that you're dishonest with the client. The fact you can overdeliver means you're making some trade-offs elsewhere, e.g. in design/production price, time, weight or somewhere else.
Overdelivering is a nice byproduct, but it shouldn't be the goal.
It has nothing to do with dishonesty. Though I have no doubt that it might be used that way, that's fraudulent and a whole different ball-game commercially.
In such a situation, why not say like it is? Just explain the client the situation, what's the risk and for what additional cost it can be lowered. And I have no doubt this is what was happening in NASA, in other words this was no "underpromise and overdeliver", this was "explain the situation and agree on trade-offs leading to the best ratio of positive outcome for the price".
I think the only situation where this approach is valid is when your client is not sophisticated enough to understand the trade-offs and make an optimal decision.
But officially the statements are always set as an "underpromise and overdeliver" thing because the clients are not sophisticated enough to understand the trade-offs.
By clients I mean politicians and the general public, who take even delays or testing-to-failure as being actual failure.
That doesn't work well when communicating indirectly, through politicians and journalists, to the public at large. It may not work when communicating to elected politicians.
That brought to mind the glass-half-full / half-empty (false) dichotomy: The engineer thinks the glass is twice as big as it needs to be (or maybe 30-40% bigger to allow for a safety margin).
That's the wrong kind of overperforming.
NASA’s got a history of their rovers overperforming like this. Between this and Webb, it’s hard to argue they’re not the best out there.
It's underpromising and overdelivering. It's a smart strategy because it's just hard to get stuff to work in space and on another planet. So it's best to underpromise.
You can bet that the internal goal is a lot higher than the public goal.