Ingenuity helicopter's flying days cut short by featureless Martian terrain It appears the bland Martian surface triggered a chain of events that left NASA's Ingenuity helicopter permanently grounded on the red planet. The helicopter's flying career came to an abrupt end earlier this year when Flight 72 was cut short, and communications were briefly lost. After re-establishing contact, it soon became …
Ginnies radio isn't powerful enough to reach earth. Currently it still has line of sight with Perseverance and Percy is relaying this communication to earth (sometimes via MRO) but once that moves out of range it's unlikely we'll ever get a signal from the little helicopter that could again.
CFIT, or 'Controlled Flight Into Terrain' is one of the most common ways of losing an aircraft (AKA 'crashing'). It could have been fitted with a radio or laser altimeter (these are highly accurate, golfers use them to determine the distance to 'holes', engineers use laser measurements to survey buildings etc.), but that would have entailed extra weight, and a power drain.
At over 14 times the successful number of flights expected, that is a class piece of engineering. And the fact that it is still sending telemetry data is awesome.
It does have an onboard laser altimeter. Here's the list of navigation sensors:
• a Bosch Sensortech BMI-160 inertial measurement unit (IMU), for measuring 3-axis accelerations at 1600 Hz and angular rates at 3200 Hz
• a Garmin Lidar-Lite-V3 laser rangefinder (LRF), for measuring distance to the ground at 50 Hz
• a downward-looking 640 × 480 grayscale camera with an Omnivision OV7251 global-shutter sensor, providing
images at 30 Hz
• a muRata SCA100T-D02 inclinometer, for measuring roll and pitch attitude prior to flight
These are commercial off-the-shelf (COTS) miniature sensors, largely developed for the cell phone and lightweight drone markets
It knew how high it was, what it couldn't figure out was the groundspeed.
The downward facing ground flow camera is used to measure orientation and ground track to correct for IMU drift. Without measurable ground flow it could only approximate its groundspeed.
The incident flight was trying to reorient itself after losing track.
Compasses don't work on Mars, and there's no GPS so it has to use visual flight rules.
Compasses don't work on Mars, and there's no GPS so it has to use visual flight rules.
So whats the alternative? In this context it seems obvious to me - you have the rover there. At the most basic fit a small radio beacon on the rover and use it as a fixed reference point. Getting more sophisticated you could have a number of beacons the probe can periodically drop on the surface (you'd need several as the rover and copter move out of range) allowing for ILS-style trig calculation of position. Or even track the copter on the rover cameras and supply real time "I have you at this aziuth and this elevation" telemetery during flight.
All seem well suited for an advanced scout support role such as this, I would be surprised if those kind of options were not under consideration. This was an experimental craft and learning exercise after all. Of course it wouldn't help for a "copter first" mission as is proposed for Titan but that is a separate issue.
Better camera resolution would probably have avoided this. The downward looking cameras aren't very high res (on purpose, higher resolution also means higher computational load to deal with that massive increase in data) which likely contributed to not being able to track any ground features.
Long term something like a Mars LORA system might be a way to handle it, but we're far from being able to deploy something like that currently. Also keep in mind that Ingenuity was a research project intended to make no more than a single handful of flights over a few sols. It was definitely not expected to last for 72 flights.
> on Mars, ....there's no GPS so it has to use visual flight rules.
Certainly we could launch low-orbit GPS. But as a short-cut we could go back to 1955 bomber training. Targets were old tanks roaming the desert (say Texas) and the bombs were light canisters full of flour. A hit would be clearly scored. (The tale I was told, the general's Cadillac Convertible was leading the line of tanks and got well-flowered. And the umpire counted it!) Anyway, one dropped plop of contrasting dye (or laser toner?) would give some lateral groundspeed. Two dots gives a little more info.
Yes, they learned an amazingly useful lesson from something they quite possibly never expected to even encounter on this "test rig/demonstrator". Now they know it's possible, they can work on something with more lift capability so they can afford to spend energy on more sensors The primary way of measuring ground speed was a low res camera, which worked really well up to the point they ended up an area akin to the Sahara. Those few limited flights were only ever expected to take place in the immediate area of the initial landing and the camera + s/w was eminently able to cope there. They were a victim of their own success and tried to fly over the "here be dragons" part of the map :-)
The whole point of missions like this is to push the envelope to see what is possible. The only way to know where the limits are is to step over them. Baby steps so you learn the most.
They did bloody well with what they did considering this was a "we've no idea whether it will actually work at all" machine.
The good old argument of 'why is X so expensive, when Y seems to work fine?'
The issue isn't that they won't work, it's that you *hope* they'll work rather than *knowing* they'll work in that environment.
Whether it's for avionics or space or military or whatever you're paying extra for the guarantee of the known reliability. If your phone gives up because it's too hot or cold or it got bounced it isn't the end of the world, if you've spent eight or 9 digits worth of dollars or people might die the requirements are a bit different.
Being able to point at something surviving the environment and say 'it worked!' doesn't change the fundamentals.
There's all sorts of situations where stuff is 99%+ likely to be fine (cars, computers, whatever) but you still spend on risk mitigation (airbags, backups, redundancy) because of the consequences when it doesn't.
For throwaway cheap missions fine, but for an important one you're going to look like a proper mug if you cost engineered the design and fate wasn't on your side and it breaks. Try explaining the brilliance of your cellphone processor choice if the radiation or temperature kills it and your mission.
risk == probability_of_bad_thing_happening * consequences_of_bad_thing_happening
1. Kudos to NASA for getting such good results from a throwaway mission.
2. "Self-driving" car makers assign far-too-low a value to the consequences -- human injury and death -- part of the equation. In other words, "zero". They only look at monetary values affecting their company: "If this fails, how likely are we to be fined or effectively sued for?", "Can we pressure or bribe judges and regulatory staff to minimize the money we'll have to pay upon our product's failure(s)?", "How much more profit can we make if we do things the fast and/or cheaper (yet lower-quality) way?"
(Helicopter icon for ... helicopter)
I read an article a while back (may have been here) that NASA originally was very conservative about the missions they had Ingenuity fly, but as they wanted to explore more interesting terrain it was flying longer missions that were deemed more risky. They knew they were increasing the chances of a catastrophic failure, but deemed it to be worth the risk as the flat featureless terrain is less interesting if you are looking for evidence of past surface water and possible life processes.
The helicopter has been commanded to continue taking pictures and sensor readings and storing them.
It has enough storage to do this for about 20 years, and is solar powered. (of course dust will be an issue)
How long will it last? And will there be any one to come retrieve that data?
After the next drive, Curiosity will be out of the line of sight for the last time.
Well for now it still can send the data home. Even when it can't what is being lost by it continuing? It is not as if it is being done at the expense of anything else. There is a chance however remote of the rover returning for some reason or reacquiring line of sight. It would be a wasted opportunity if that did occur and the data hadn't been collected.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
