Author Topic: How Can AI Be Used for Space Applications?  (Read 90665 times)

Offline InterestedEngineer

  • Senior Member
  • *****
  • Posts: 2699
  • Seattle
  • Liked: 2090
  • Likes Given: 3406
Re: How Can AI Be Used for Space Applications?
« Reply #400 on: 08/01/2024 03:33 pm »
AI as a search engine for large technical manuals on the space station:

https://spacenews.com/booz-allen-deploys-advanced-language-model-in-space/

Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #401 on: 08/01/2024 07:19 pm »
Dr Becky - How we use AI in Astrophysics:



It seems overall AI hasn’t revolutionised astrophysics unlike something like Alphafold has done, instead it’s just sped things up somewhat.
« Last Edit: 08/01/2024 07:40 pm by Star One »

Offline edzieba

  • Virtual Realist
  • Senior Member
  • *****
  • Posts: 6782
  • United Kingdom
  • Liked: 10389
  • Likes Given: 46
Re: How Can AI Be Used for Space Applications?
« Reply #402 on: 08/02/2024 09:54 am »
It also helped that 'AI' (in its current zeitgeist meaning use of MLNNs for image processing) has been in use for astronomy - and elsewhere - since well before the current 'deep learning' boom started. Wide availability of hardware to support NN acceleration has made the existing techniques cheaper and faster, but the techniques themselves have not really been new.

Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #403 on: 08/02/2024 04:32 pm »
It also helped that 'AI' (in its current zeitgeist meaning use of MLNNs for image processing) has been in use for astronomy - and elsewhere - since well before the current 'deep learning' boom started. Wide availability of hardware to support NN acceleration has made the existing techniques cheaper and faster, but the techniques themselves have not really been new.
It says in the video that astrophysicists have to create bespoke machine learning systems each time they are trying to investigate a particular matter because of the data sets they work with tending to be unique and large.

Offline marcin77

  • Member
  • Posts: 1
  • Liked: 0
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #404 on: 08/03/2024 01:01 pm »
Artificial intelligence  is playing an increasingly significant role in various fields, including space exploration. How exactly is AI utilized in space research? What are the most promising applications of AI in space missions, data analysis and other aspects of space exploration? I’ve come across some general information on a few websites, but I’m looking for more specific details. What are the most innovative and groundbreaking uses of AI in space? Can you provide examples of recent AI applications in space missions or projects?

Offline zubenelgenubi

  • Global Moderator
  • Senior Member
  • *****
  • Posts: 12592
  • Arc to Arcturus, then Spike to Spica
  • Sometimes it feels like Trantor in the time of Hari Seldon
  • Liked: 8487
  • Likes Given: 83581
Re: How Can AI Be Used for Space Applications?
« Reply #405 on: 08/03/2024 02:01 pm »
[questions]
I merged your post to the thread that may answer your questions.

Welcome to the forum!
Support your local planetarium! (COVID-panic and forward: Now more than ever.) My current avatar is saying "i wants to go uppies!" Yes, there are God-given rights. Do you wish to gainsay the Declaration of Independence?

Offline LMT

  • Lake Matthew Team
  • Senior Member
  • *****
  • Posts: 2577
    • Lake Matthew
  • Liked: 432
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #406 on: 08/09/2024 08:07 pm »
In LEO, many autonomous or telerobotic operations seem feasible near-term. 

With tiny inboard fans, an AI bot could move freely within a station.  Step into a tug's foot restraints, and it could move freely outside the station.

Business cases?

Cross-post on station assembly robotics.

Optimus hands at 5:37.


« Last Edit: 08/09/2024 08:21 pm by LMT »

Offline LMT

  • Lake Matthew Team
  • Senior Member
  • *****
  • Posts: 2577
    • Lake Matthew
  • Liked: 432
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #407 on: 08/12/2024 04:41 pm »
Xin, M., You, Z., Zhang, Z., Jiang, T., Xu, T., Liang, H., Ge, G., Ji, Y., Mo, S. and Cheng, J., 2024. We Choose to Go to Space: Agent-driven Human and Multi-Robot Collaboration in Microgravity. arXiv preprint arXiv:2402.14299.



Offline JulesVerneATV

  • Regular
  • Full Member
  • **
  • Posts: 217
  • Liked: 25
  • Likes Given: 4
Re: How Can AI Be Used for Space Applications?
« Reply #408 on: 08/24/2024 09:09 am »
Shape-shifting robot inspired by insect swarms and tree roots is teaching itself to mark contamination zones
https://www.spacedaily.com/reports/Shape_shifting_WVU_robot_inspired_by_insect_swarms_and_tree_roots_is_teaching_itself_to_mark_contamination_zones_999.html

Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #409 on: 08/28/2024 08:32 am »
From what I’ve read it seems to be believed  by a number of experts in this area that many of the technical difficulties will be solved in the open source arena, it’s why FB made their AI open source I expect. The old saying being many heads are better than a few. See below.

Quote
Nous Research turned heads earlier this month with the release of its permissive, open-source Llama 3.1 variant Hermes 3.

Now, the small research team dedicated to making “personalized, unrestricted AI” models has announced another seemingly massive breakthrough: DisTrO (Distributed Training Over-the-Internet), a new optimizer that reduces the amount of information that must be sent between various GPUs (graphics processing units) during each step of training an AI model.
But Nous Research, whose whole approach is essentially the opposite — making the most powerful and capable AI it can on the cheap, openly, freely, for anyone to use and customize as they see fit without many guardrails — has found an alternative.
Yet, the authors also say that “our preliminary tests indicate that it is possible to get a bandwidth requirements reduction of up to 1000x to 3000x during the pre-training,” phase of LLMs, and “for post-training and fine-tuning, we can achieve up to 10000x without any noticeable degradation in loss.”



Quote
They further hypothesize that the research, while initially conducted on LLMs, could be used to train large diffusion models (LDMs) as well: think the Stable Diffusion open source image generation model and popular image generation services derived from it such as Midjourney.
To be clear: DisTrO still relies on GPUs — only instead of clustering them all together in the same location, now they can be spread out across the world and communicate over the consumer internet.

Specifically, DisTrO was evaluated using 32x H100 GPUs, operating under the Distributed Data Parallelism (DDP) strategy, where each GPU had the entire model loaded in VRAM.
By reducing the need for high-speed interconnects DisTrO could enable collaborative model training across decentralized networks, even with participants using consumer-grade internet connections.



Quote
The report also explores the implications of DisTrO for various applications, including federated learning and decentralized training.

Additionally, DisTrO’s efficiency could help mitigate the environmental impact of AI training by optimizing the use of existing infrastructure and reducing the need for massive data centers.

https://venturebeat.com/ai/this-could-change-everything-nous-research-unveils-new-tool-to-train-powerful-ai-models-with-10000x-efficiency/

Paper linked to at end of the article.

Online Vultur

  • Senior Member
  • *****
  • Posts: 2189
  • Liked: 926
  • Likes Given: 184
Re: How Can AI Be Used for Space Applications?
« Reply #410 on: 08/28/2024 01:38 pm »
I remain skeptical (in terms of space or otherwise technical relevance). No matter how cheaply you can train them, LLMs are just too unreliable/unmoored from reality to be used for anything technical, safety-critical, etc. They're great for generating corporate boilerplate or form letters quickly, but...

"AI" that would be useful for advanced space probes, ISRU mining on Mars before human crews arrive, etc. etc. won't come from the LLM/LDM type tech at all.

Offline LMT

  • Lake Matthew Team
  • Senior Member
  • *****
  • Posts: 2577
    • Lake Matthew
  • Liked: 432
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #411 on: 08/28/2024 02:04 pm »
I remain skeptical (in terms of space or otherwise technical relevance). No matter how cheaply you can train them, LLMs are just too unreliable/unmoored from reality to be used for anything technical, safety-critical, etc. They're great for generating corporate boilerplate or form letters quickly, but...

"AI" that would be useful for advanced space probes, ISRU mining on Mars before human crews arrive, etc. etc. won't come from the LLM/LDM type tech at all.

Quote from: Xin et al. 2024
The capability for collaborative task planning originates from the foundation model OpenAI (2023), which directly outputs the collaboration graph in JSON format guided by prompt engineering Mo and Xin (2023).
« Last Edit: 08/28/2024 02:05 pm by LMT »

Online Vultur

  • Senior Member
  • *****
  • Posts: 2189
  • Liked: 926
  • Likes Given: 184
Re: How Can AI Be Used for Space Applications?
« Reply #412 on: 08/28/2024 08:56 pm »
Planning is a conceivable use -- though if I am reading the table you provided correctly, the AI still perforned worse than human experts, just close-ish. Sure, there is room for improvement, but given things like model collapse, how much? (Once LLMs have ingested all available information, and started to get polluted by LLM-generated stuff in their inputs...)

Also, unless the hallucination problem can be fixed (and it is likely not possible as it's inherent to how LLMs work) they'll never be trustworthy for anything important.

My extreme skepticism is for things like automating probes/rovers/etc (so they can respond faster than light-lagged communications to and from controllers on Earth), ISRU, and so on - tasks based on interaction with a physical environment. LLMs are language models, based on manipulating strings of characters/words/tokens. They are terrible choices for AI to do physical things.

Offline LMT

  • Lake Matthew Team
  • Senior Member
  • *****
  • Posts: 2577
    • Lake Matthew
  • Liked: 432
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #413 on: 08/28/2024 09:54 pm »
My extreme skepticism is for... tasks based on interaction with a physical environment.

And you just saw that.   ::)

Foundation models have broad application, of course.  SpaceAgents-1 multi-robot control directed robots with simple manipulators, in microgravity sim.  Now extend control to, say, a free-flying Optimus 1 2 with its dexterous hands.  New use cases suggest themselves.


« Last Edit: 08/28/2024 11:04 pm by LMT »

Offline LMT

  • Lake Matthew Team
  • Senior Member
  • *****
  • Posts: 2577
    • Lake Matthew
  • Liked: 432
  • Likes Given: 0
Re: How Can AI Be Used for Space Applications?
« Reply #414 on: 09/16/2024 01:26 pm »
SPAICE 2024, Sept. 17-19, with many relevant poster sessions, e.g.:

Quote
9/19/24
Trajectory Optimization with Reinforcement Learning-driven Control of Multi-Arm Robots in On-Orbit Servicing Operations

Celia Redondo-Verdú & Jorge Pomares (University of Alicante)

Intelligent motion planning and control systems are crucial in managing the inherently unpredictable and non-deterministic conditions of space. Trajectory optimization (TO) has been a key method in space robotics for precision control and guidance. However, the effectiveness of TO is limited by its reliance on accurate robotic and environmental models. Reinforcement Learning (RL) has emerged as a promising alternative, enhancing robustness and adaptability in control systems. Unlike TO, RL does not depend on predefined models but learns effective control policies through simulated interactions. This model-free approach has shown potential in handling complex control tasks and adapting to environmental uncertainties and perturbations. This paper proposes an integrated TO and RL approach for enhanced path planning and control of a four-arm robot designed for on-orbit servicing tasks. By combining the predictive power of TO with the adaptive capabilities of RL, this approach aims to optimize the robot’s operational effectiveness. The integration seeks to leverage the detailed pre-planned motion profiles provided by TO, while incorporating the flexibility and resilience of RL-based control strategies to accommodate real-time operational dynamics and uncertainties. The results presented demonstrate how this hybrid approach can significantly improve the robot’s trajectory tracking performance and operational adaptability.
« Last Edit: 09/16/2024 01:28 pm by LMT »

Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #415 on: 09/25/2024 07:51 pm »
Quote
I am going to walk you through some advanced AI magic and bring you up-to-speed on state-of-the-art boundary-pushing when it comes to generative AI. It is a bit of a mystery story too.

In this column, I will showcase something that is on the outskirts of everyday generative AI and primarily experimental in advanced AI labs. It is an approach that leverages multiple chain-of-thought processing and incorporates AI-based meta-reasoning. Some believe that this might be an essential ingredient or secret sauce of the new o1 generative AI model, and, by the way, the future for leading-edge generative AI.

https://www.forbes.com/sites/lanceeliot/2024/09/19/speculation-brims-that-openai-o1-leverages-multiple-chain-of-thought-and-meta-reasoning/

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39442
  • Minnesota
  • Liked: 25544
  • Likes Given: 12223
Re: How Can AI Be Used for Space Applications?
« Reply #416 on: 09/29/2024 02:24 am »
BTW, NASA is in the midst of an agency-wide push (from HQ?) for AI (as well as what to be careful about with respect to AI). Some big agency directive I think came down a few months ago.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #417 on: 10/08/2024 11:48 pm »
Quote
The Nobel Prize in Physics has been awarded to two scientists, Geoffrey Hinton and John Hopfield, for their work on machine learning.
British-Canadian Professor Hinton is sometimes referred to as the "Godfather of AI" and said he was flabbergasted.
He resigned from Google in 2023, and has warned about the dangers of machines that could outsmart humans.

https://www.bbc.co.uk/news/articles/c62r02z75jyo.amp

Here’s the related interview:


Offline Star One

  • Senior Member
  • *****
  • Posts: 14328
  • UK
  • Liked: 4099
  • Likes Given: 220
Re: How Can AI Be Used for Space Applications?
« Reply #418 on: 10/17/2024 02:09 pm »
The use of small modular reactors to power AI is obviously relevant to spaceflight.

Quote
On Monday, Google announced a landmark agreement with nuclear startup Kairos Power to purchase energy produced by seven yet-to-be-built small modular nuclear reactors. The companies claim the deal aims to add upwards of 500 megawatts "of new 24/7 carbon-free power to US electricity grids" — that is, over a decade from now when Kairos promises the reactors will be built.



Quote
In the announcement, Google and Kairos claim that the first of the modular nuclear power terminals will be up and running by 2030, with all modules completed by 2035.

https://futurism.com/the-byte/google-nuclear-power

Google press release:

https://blog.google/outreach-initiatives/sustainability/google-kairos-power-nuclear-energy-agreement/

Offline Asteroza

  • Senior Member
  • *****
  • Posts: 2998
  • Liked: 1156
  • Likes Given: 33
Re: How Can AI Be Used for Space Applications?
« Reply #419 on: 10/17/2024 10:59 pm »
The use of small modular reactors to power AI is obviously relevant to spaceflight.

Quote
On Monday, Google announced a landmark agreement with nuclear startup Kairos Power to purchase energy produced by seven yet-to-be-built small modular nuclear reactors. The companies claim the deal aims to add upwards of 500 megawatts "of new 24/7 carbon-free power to US electricity grids" — that is, over a decade from now when Kairos promises the reactors will be built.



Quote
In the announcement, Google and Kairos claim that the first of the modular nuclear power terminals will be up and running by 2030, with all modules completed by 2035.

https://futurism.com/the-byte/google-nuclear-power

Google press release:

https://blog.google/outreach-initiatives/sustainability/google-kairos-power-nuclear-energy-agreement/

Amazon is now backing X-energy, another nuclear startup

https://www.aboutamazon.com/news/sustainability/amazon-nuclear-small-modular-reactor-net-carbon-zero

Tags:
 

Advertisement NovaTech
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0