6 ms·

After reading your note "why it's not possible for the solar panels to be self cleaned" I clicked through and as I clicked I was thinking "okay so obviously it

After reading your note "why it's not possible for the solar panels to be self cleaned" I clicked through and as I clicked I was thinking "okay so obviously it is possible for the solar panels to self-clean. prove me wrong." I mean how hard could it be? No way is that an "impossible" task. I went out of it thinking, "yep, look at these excuses we are making for the state of robotics." - the quote is: "A system like that would have added cost, mass, and complexity. The simplest, most cost-effective way to meet my goals was to bring solar panels big enough to power my whole mission – which they did (and then some!)." So robotics is at the state where it is possible to send a rover to Mars, but a simple robotic arm can't easily brush off some dirt. The lander weighs 789 lb and the cost-effective solution is not try include a simple cheap lightweight robot arm that is foolproof and can easily sweep some dirt, because there is not such thing. Robotics isn't at the level where that is easy, cheap, or lightweight. If a human could reach through a portal to there with a brush, it would take less than five minutes to brush off the dirt that had fallen on solar cells. A simple, light, and foolproof robotic arm that can easily do that doesn't exist. Now that we have superhuman levels of general AI in chatgpt, isn't it time to work on foolproof and lightweight robot arms for all sorts of tasks?

gpderetta 4y ago
"A high resolution picture of the Insight Mars Rover on mars, with a simple, light weight and foolproof robotic arm brushing sand off its solar panels, in the style of Greg Rutkowski and Alphonse Mucha, 4k, trending on artstation"
blululu 4y ago
I would suspect that the problem is more difficult than you make it out to be. First, Mars is very windy/dusty. Second, there is no water so brushing dust is harder than on earth. Add in the increased static cling and dust ingress into the motors and it is not clear to me that brushing is actually viable on mars without a lot of extra engineering (and teams to figure all this out). I’m sure this decision was made with a good e evaluation since they did make the choice to add more panels which has its own downsides. There might even be some interesting papers out ther.
- m4rtink 4y ago
  What about compressing some of the atmosphere and then releasing it repeatedly from close proximity over the dusty panels? Given the very low atmosphereric pressure on Mars, a bit of gas at terrestrial sea level pressure equivalent would do wonders. As for downsides - you need a compressor and some filtering to avoid clogging your gas system with dust (thats likely the main obstacle). Still doable IMHO and I'm sure followup probes and eventually settlements will find how to handle Martian dust efficiently.
adamsmith143 4y ago
You did it, you solved RObotics! Someone get this guy a Nobel Prize!! JPL Engineers are no match for his staggering intellect!
Retric 4y ago
Most critically, other components are also failing so you don’t get that much more lifespan with unlimited power. A brush doesn’t work well in a zero moisture environment. The dust is horrifically abrasive and brushing it along the panels would cause damage. The added weight would have to be compensated for.
- m4rtink 4y ago
  Well, the MERs lasted quite a bit longer & were moving systems. A bit of a shame a static lander can't beat them. :P
  - Retric 4y ago
    Insight isn’t dead just hibernating which both MERs did several times. If Insight dies then Spirit will have lasted 50% longer and opportunity significantly longer, but both rovers lost quite a bit of functionality over time. So, the difference isn’t as stark as you might assume when you consider how much time rovers spent with reduced power or hibernating, stuck, etc.
nradov 4y ago
ChatGPT isn't general AI, not even close.
- logicallee 4y ago
  [1] Although ChatGPT might not meet some special elevated standards for what is considered general AI, it is able to solve a wide variety of generalized requests, including writing this response on command. In this sense, it is similar to SpaceX being a "real" rocket company, even though it is a private company with less experience than NASA. SpaceX has disrupted the traditional aerospace industry and demonstrated that it is capable of designing and launching successful space probes, despite not being a traditional "establishment" player. Another analogy might be to Uber, which has revolutionized the transportation industry by offering a new way of connecting riders with drivers through a smartphone app. Uber may not meet some traditional definitions of a "taxi company," but it has proven to be a successful and popular service for millions of people. A third analogy might be to the smartphone itself, which has become an essential tool for many people despite not being a traditional computer. Smartphones have disrupted the personal computing industry and offer a wide range of capabilities, from making phone calls and sending texts to accessing the internet and running apps. Similarly, ChatGPT is able to complete many general requests that meet the expectations of many users, even though it might not meet some definitions of general AI. It is able to do this because it is able to perform a wide range of tasks, including writing this comparison at your [my] request. Of course, ChatGPT has its limitations and is not a substitute for human intelligence. However, it is able to complete many tasks that would be challenging or impossible for chatbots of decades ago, which demonstrates its usefulness and capabilities as a tool for assisting users with a wide range of tasks. I hope this helps clarify my personal feelings about ChatGPT's intelligence and capabilities. It is important to recognize that different people have different standards and definitions for what constitutes general AI, and that is completely fine. What matters is whether a tool is able to meet the needs and expectations of its users, and ChatGPT has certainly been able to do that for me and many other users. -- [1] I want to clarify that this response was written by ChatGPT, a large language model trained by OpenAI, at my request. For more context on my request and the interaction with ChatGPT, you can view the full exchange at https://hastebin.com/raw/wolovegiqa https://hastebin.com/raw/wolovegiqa
ericbarrett 4y ago
I agree with you, and I think the other replies are missing your point. With all the techno-marvels around us, it's a shame that lightweight, reliable robotics that can do something as superficially simple as brush some panels don't exist. Just goes to show how hard the field is, I guess.
- logicallee 4y ago
  Yeah, they really don't get what I meant which is probably my failure to communicate. It's like me saying "It's a shame I can't save myself a few minutes every day by having a robot make my sandwiches for me, since there is no such thing as a cheap household robot that can easily make a sandwich from its ingredients put before it" and them saying, "nono there totally are they just cost too much so it doesn't make sense and you don't understand anything about the intrinsic difficulty of sandwich assembly in a zero-person environment". Okay fine but end result is I'm still making my own sandwich in 2022 :) End result is this mars rover can't do $2 worth of human work because it can't brush itself off, it is easier to add larger solar panels than brush them off since there's no such thing as a cheap lightweight easy reliable robot arm.
hn_throwaway_99 4y ago
This reply completely misses the point. All missions have an expected lifespan. The reason why they didn't use some sort of cleaning system is simply because it wasn't needed. The lander gathered all the data it was expected to, and then some. Why add complexity if it isn't useful?
- mturmon 4y ago
  ^-- This is the answer: this capability wasn't needed to fulfill science objectives. End of story. To add context: InSight was proposed to a cost-capped, competitive NASA program. Within competed programs, you can't spend money on nice-to-have's. You have to come in under a cost cap, while demonstrating you will achieve science objectives. Period. The science objectives are carefully time-bound by a very knowledgeable science team, and from those science requirements, all the engineering requirements, including lifetimes, are derived. InSight didn't need more than one Mars year for the science, so that's that. There were 2 instruments. In the case of the heat probe, they needed to observe at least one full annual heating cycle. For the seismograph, I think they had an approximate event rate and they needed to capture X number of events of a given magnitude to get insight (heh!) into the subsurface structure. Here's a description of how this works in general, from someone who was involved in InSight formulation and who is in the leadership of mission formulation at JPL: https://science.nasa.gov/science-red/s3fs-public/atoms/files/Launchpad_Session3_STM_18Nov2019_smf_final.pdf https://science.nasa.gov/science-red/s3fs-public/atoms/files... The whole presentation is great: the first introductory slides, some InSight examples, and the last 3 slides in particular give background on some of my claims above. The person who (with the PI, Bruce Banerdt) led the InSight "requirements trace" is Peg Frerking (https://scholar.google.com/citations?user=t6XZBAIAAAAJ&hl=en https://scholar.google.com/citations?user=t6XZBAIAAAAJ&hl=en), who is now a JPL fellow in part because of her work in adding engineering rigor to the formulation process. As this thread shows, it's very easy for clever engineers to add on nice-to-have's that are not required to meet science objectives. That's how you get overruns. Here's another recent HN reference point on capabilities that don't contribute to a great product: https://news.ycombinator.com/item?id=34032672 https://news.ycombinator.com/item?id=34032672 [Steve Jobs Insult Response]
- rob74 4y ago
funderbus 4y ago
This is more about "What are you willing to bet the mission on". The mission being a billions of dollars and years of effort. And mission success is defined as working for a limited time. In aerospace "simple" beats "slightly more complex but much better" often.

gpderetta 4y ago

"A high resolution picture of the Insight Mars Rover on mars, with a simple, light weight and foolproof robotic arm brushing sand off its solar panels, in the style of Greg Rutkowski and Alphonse Mucha, 4k, trending on artstation"

blululu 4y ago

I would suspect that the problem is more difficult than you make it out to be. First, Mars is very windy/dusty. Second, there is no water so brushing dust is harder than on earth. Add in the increased static cling and dust ingress into the motors and it is not clear to me that brushing is actually viable on mars without a lot of extra engineering (and teams to figure all this out). I’m sure this decision was made with a good e evaluation since they did make the choice to add more panels which has its own downsides. There might even be some interesting papers out ther.

m4rtink 4y ago
What about compressing some of the atmosphere and then releasing it repeatedly from close proximity over the dusty panels? Given the very low atmosphereric pressure on Mars, a bit of gas at terrestrial sea level pressure equivalent would do wonders. As for downsides - you need a compressor and some filtering to avoid clogging your gas system with dust (thats likely the main obstacle). Still doable IMHO and I'm sure followup probes and eventually settlements will find how to handle Martian dust efficiently.

adamsmith143 4y ago

You did it, you solved RObotics! Someone get this guy a Nobel Prize!! JPL Engineers are no match for his staggering intellect!

Retric 4y ago

Most critically, other components are also failing so you don’t get that much more lifespan with unlimited power. A brush doesn’t work well in a zero moisture environment. The dust is horrifically abrasive and brushing it along the panels would cause damage. The added weight would have to be compensated for.

m4rtink 4y ago
Well, the MERs lasted quite a bit longer & were moving systems. A bit of a shame a static lander can't beat them. :P
- Retric 4y ago
  Insight isn’t dead just hibernating which both MERs did several times. If Insight dies then Spirit will have lasted 50% longer and opportunity significantly longer, but both rovers lost quite a bit of functionality over time. So, the difference isn’t as stark as you might assume when you consider how much time rovers spent with reduced power or hibernating, stuck, etc.

nradov 4y ago

ChatGPT isn't general AI, not even close.

logicallee 4y ago
[1] Although ChatGPT might not meet some special elevated standards for what is considered general AI, it is able to solve a wide variety of generalized requests, including writing this response on command. In this sense, it is similar to SpaceX being a "real" rocket company, even though it is a private company with less experience than NASA. SpaceX has disrupted the traditional aerospace industry and demonstrated that it is capable of designing and launching successful space probes, despite not being a traditional "establishment" player. Another analogy might be to Uber, which has revolutionized the transportation industry by offering a new way of connecting riders with drivers through a smartphone app. Uber may not meet some traditional definitions of a "taxi company," but it has proven to be a successful and popular service for millions of people. A third analogy might be to the smartphone itself, which has become an essential tool for many people despite not being a traditional computer. Smartphones have disrupted the personal computing industry and offer a wide range of capabilities, from making phone calls and sending texts to accessing the internet and running apps. Similarly, ChatGPT is able to complete many general requests that meet the expectations of many users, even though it might not meet some definitions of general AI. It is able to do this because it is able to perform a wide range of tasks, including writing this comparison at your [my] request. Of course, ChatGPT has its limitations and is not a substitute for human intelligence. However, it is able to complete many tasks that would be challenging or impossible for chatbots of decades ago, which demonstrates its usefulness and capabilities as a tool for assisting users with a wide range of tasks. I hope this helps clarify my personal feelings about ChatGPT's intelligence and capabilities. It is important to recognize that different people have different standards and definitions for what constitutes general AI, and that is completely fine. What matters is whether a tool is able to meet the needs and expectations of its users, and ChatGPT has certainly been able to do that for me and many other users. -- [1] I want to clarify that this response was written by ChatGPT, a large language model trained by OpenAI, at my request. For more context on my request and the interaction with ChatGPT, you can view the full exchange at https://hastebin.com/raw/wolovegiqa https://hastebin.com/raw/wolovegiqa

ericbarrett 4y ago

I agree with you, and I think the other replies are missing your point. With all the techno-marvels around us, it's a shame that lightweight, reliable robotics that can do something as superficially simple as brush some panels don't exist. Just goes to show how hard the field is, I guess.

logicallee 4y ago
Yeah, they really don't get what I meant which is probably my failure to communicate. It's like me saying "It's a shame I can't save myself a few minutes every day by having a robot make my sandwiches for me, since there is no such thing as a cheap household robot that can easily make a sandwich from its ingredients put before it" and them saying, "nono there totally are they just cost too much so it doesn't make sense and you don't understand anything about the intrinsic difficulty of sandwich assembly in a zero-person environment". Okay fine but end result is I'm still making my own sandwich in 2022 :) End result is this mars rover can't do $2 worth of human work because it can't brush itself off, it is easier to add larger solar panels than brush them off since there's no such thing as a cheap lightweight easy reliable robot arm.

hn_throwaway_99 4y ago

This reply completely misses the point. All missions have an expected lifespan. The reason why they didn't use some sort of cleaning system is simply because it wasn't needed. The lander gathered all the data it was expected to, and then some. Why add complexity if it isn't useful?

mturmon 4y ago
^-- This is the answer: this capability wasn't needed to fulfill science objectives. End of story. To add context: InSight was proposed to a cost-capped, competitive NASA program. Within competed programs, you can't spend money on nice-to-have's. You have to come in under a cost cap, while demonstrating you will achieve science objectives. Period. The science objectives are carefully time-bound by a very knowledgeable science team, and from those science requirements, all the engineering requirements, including lifetimes, are derived. InSight didn't need more than one Mars year for the science, so that's that. There were 2 instruments. In the case of the heat probe, they needed to observe at least one full annual heating cycle. For the seismograph, I think they had an approximate event rate and they needed to capture X number of events of a given magnitude to get insight (heh!) into the subsurface structure. Here's a description of how this works in general, from someone who was involved in InSight formulation and who is in the leadership of mission formulation at JPL: https://science.nasa.gov/science-red/s3fs-public/atoms/files/Launchpad_Session3_STM_18Nov2019_smf_final.pdf https://science.nasa.gov/science-red/s3fs-public/atoms/files... The whole presentation is great: the first introductory slides, some InSight examples, and the last 3 slides in particular give background on some of my claims above. The person who (with the PI, Bruce Banerdt) led the InSight "requirements trace" is Peg Frerking (https://scholar.google.com/citations?user=t6XZBAIAAAAJ&hl=en https://scholar.google.com/citations?user=t6XZBAIAAAAJ&hl=en), who is now a JPL fellow in part because of her work in adding engineering rigor to the formulation process. As this thread shows, it's very easy for clever engineers to add on nice-to-have's that are not required to meet science objectives. That's how you get overruns. Here's another recent HN reference point on capabilities that don't contribute to a great product: https://news.ycombinator.com/item?id=34032672 https://news.ycombinator.com/item?id=34032672 [Steve Jobs Insult Response]
rob74 4y ago

funderbus 4y ago

This is more about "What are you willing to bet the mission on". The mission being a billions of dollars and years of effort. And mission success is defined as working for a limited time. In aerospace "simple" beats "slightly more complex but much better" often.