... you'll need a battery with that!
BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Frangipani
Top 20
Dr Kenneth Östberg presented our poster on GRAIN – Radiation-Tolerant Edge… | Frontgrade Gaisler
Dr Kenneth Östberg presented our poster on GRAIN – Radiation-Tolerant Edge AI at the RISC-V in Space Workshop 2025 earlier this week. The GRAIN line of space computing products includes the newly announced rad-hard GR801 neuromorphic processor based on the BrainChip Akida 1.0 technology. Read...www.linkedin.com
This poster titled GRAIN - Radiation-Tolerant Edge AI, presented by Kenneth Östberg, one of its two co-authors (the other being Daniel Andersson), during the “RISC-V in Space” workshop in Gothenburg on Thursday…
View attachment 81573
View attachment 81577
…finally reveals what NEURAVIS stands for - the name of that R&T project, which ESA awarded to the five consortium partners Airbus Toulouse, Airbus Ottobrunn, BrainChip, Frontgrade Gaisler and Neurobus in mid-2024 (see the July 2024 LinkedIn post by Airbus Space Project Manager Jérémy Lebreton below):
Neuromorphic Evaluation of Ultra-low-power Rad-hard Acceleration for Vision Inferences in Space.
The poster also provides more information with regards to the use cases currently being explored in the NEURAVIS project, although I’m afraid I couldn’t decipher everything due to the small print - maybe someone with eagle eyes or a magic tool to blow up the photo and unblur the small print can add in resp. correct what I’ve gathered so far:
1. Moon landing
Use Case #1: Vision-Based Navigation for Lunar Lander
Also see Alf Kuchenbuch’s recent comment on Argonaut, ESA’s lunar lander programme:
https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-452257
View attachment 81579
2. Debris detection/collect (?)
Use Case #2: Monitoring (?) Building Block for In-orbit Maintenance
3. Docking
4. Object ? (looks like “simulation”, but appears to be a longer word?)
In addition, the poster lists four “Application scenarios” for GRAIN’s Radiation-Tolerant Edge AI:
1. Remote Terminal Unit
2. Stand-alone Controller
3. Near-edge processing unit
4. Auxiliary data-processing module
Lots of small print to decipher here as well!
Thank you, BrainChip team, for sharing the news. At Airbus, we are very… | Jérémy Lebreton
Thank you, BrainChip team, for sharing the news. At Airbus, we are very glad to be starting this new ESA R&T project called NEURAVIS with BrainChip , Neurobus and Frontgrade Gaisler. Our goal is to bring the disruptive technology of neuromorphic computing on an achievable path to the space...
View attachment 81578
If I understand the above post correctly, we have yet to hear about what suggestion the NEURAVIS proposal has for AKD1500 (“BrainChip is proud that Airbus selected Akida for both COTS chips and IP in their proposal. ESA awarded the Airbus “NEURAVIS” proposal, including Akida in the Akida 1500 chip and on an FPGA together with Frontgrade Gaisler’s NOEL-V processor.”).
Whereas the underlined appears to refer to Frontgrade Gaisler’s newly revealed GR801 SoC that will incorporate Akida 1.0 IP - greatly benefitting the work of Airbus Toulouse computer vision experts such as Jérémy Lebreton (project lead) and Roland Brochard, as can be inferred from the GRAIN poster’s four listed use cases - there has to be another specific proposal by Airbus how to utilise our COTS chip AKD1500, then.
So I presume Airbus Ottobrunn and Neurobus might be the consortium partners currently collaborating on that second part of the NEURAVIS proposal?
View attachment 81582 View attachment 81583
View attachment 81584
Here is a close-up of the GRAIN - Radiation-Tolerant Edge AI poster Kenneth Östberg presented at the “RISC-V in Space” workshop last month - so no more deciphering of the small print needed
:As I already mentioned the other day
, it finally reveals what NEURAVIS stands for: Neuromorphic Evaluation of Ultra-low-power Rad-hard Acceleration for Vision Inferences in Space.Attachments
Last edited:
MDhere
Top 20
His post WAS removed thank goodness, but now its back here lol
MDhere
Top 20
Moving back onto a better subject called Brainchip. The SAB of Brainchip Andre Van Schaik, I am intrigued by this man and also Gregory Cohen. I am hoping that one day our technology will be in amateur telescopes which these two are interested in as well.
I would love to buy a telescope and know exactly what I'm looking and at the same time track the event base image and still know I am looking at the same thing and then something else comes into the visual and it's something new. Would be super cool and educational. Will be keeping an eye on Andre in Manchester, where I think Brainchip may pay a visit there in July?
I would love to buy a telescope and know exactly what I'm looking and at the same time track the event base image and still know I am looking at the same thing and then something else comes into the visual and it's something new. Would be super cool and educational. Will be keeping an eye on Andre in Manchester, where I think Brainchip may pay a visit there in July?
MDhere
Top 20
Bernd Westhoff, love this man at Renesas.
www.linkedin.com
Just Launched: The Renesas RA0E2 Group – Power-Efficient. Cost-Optimized. Scalable.
We’re excited to introduce the RA0E2, the newest member of the Renesas RA Family of Arm® Cortex®-M23 MCUs, designed to meet the needs of today’s cost-sensitive and power-conscious embedded designs. With best-in-class low power consumption, a wide operating voltage and temperature range, and integrat
Fullmoonfever
Top 20
Frangipani
Top 20
You can find the presentation slides from Douglas McLelland’s talk that @Fullmoonfever shared on the conference material webpage:
Think Frontgrade got month wrong .. should be April. What is it with tech companies and simple errors
The RISC-V in Space Workshop, organized by the European Space Agency in collaboration with RISC-V International, was held on May 2–3, 2025, in Gothenburg, Sweden.
In addition, they managed to upload the wrong paper (totally unrelated to Akida) with regards to Douglas McLelland’s presentation:
Conference Material - RISC-V in Space Workshop 2025
www.risc-v.space
It should of course have been this one:
Low-power Ship Detection in Satellite Images Using Neuromorphic Hardware by the presenter himself and his co-author Gregor Lenz, who was the CTO of Neurobus at the time of publication in June 2024.
As I shared the other day, Gregor Lenz recently stepped down from his position to join Paddington Robotics in London, where he lives: https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-456183 and https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-456824).
Good to know somebody with hands-on experience of Akida is onboard.
Slightly more (but still very little) information on their website now:
https://paddington-robotics.com/
Here is an article published earlier today that explains what is meant by Embodied AI:
After the Humanoids: The Real Future of Embodied AI
UK’s Embodied AI revolution redefines robotics beyond humanoids. Paddington Robotics and Imagine Health tackle hardware chaos and edge cases, driving practical innovation with £1.3B support. Discover the future now!
newsroom.stelia.ai
After the Humanoids: The Real Future of Embodied AI
UK’s Embodied AI revolution redefines robotics beyond humanoids. Paddington Robotics and Imagine Health tackle hardware chaos and edge cases, driving practical innovation with £1.3B support. Discover the future now!
April 24, 2025
The UK is a global leader in Embodied AI (EAI), where AI powers physical systems to learn and adapt in real-world settings, from self-driving cars to healthcare diagnostics. With over 3,000 AI firms and £1.3 billion in government backing, the UK is shaping the future of robotics beyond sci-fi fantasies. I recently attended a TechAI panel featuring Zehan Wang (founder of Paddington Robotics) and Mohan Mahadevan (CEO Founder at Imagine Health). Their insights revealed where EAI is really heading, and it’s not about the humanoids.
About the Companies
Paddington RoboticsFounded in 2024 in London, Paddington Robotics is a startup tackling AI for general-purpose robotics. Led by Zehan Wang, a former Twitter AI executive with a PhD from Imperial College London, the company aims to enable widespread robotics adoption by solving real-world data challenges. With $2.26 million in funding from 7percent Ventures, their work on adaptive, non-humanoid systems aligns with EAI’s focus on practical, environment-driven intelligence, starting with applications in spaces like supermarkets.
Imagine Health
Imagine Health, a London-based startup founded in 2025, automates ultrasound diagnostics using embodied AI and robotic arms guided by computer vision. Co-Founder and CEO Mohan Mahadevan, with 20+ years in AI and robotics from Amazon and Tractable, drives its mission to deliver safe, scalable healthcare solutions, reducing practitioner strain.
EAI and Robotics – What’s the Difference?
Embodied AI (EAI) and robotics are related but distinct. EAI focuses on AI systems integrated into physical forms that learn and adapt through environmental interaction, emphasizing intelligent, adaptive behavior (e.g., Wayve’s self-driving cars). Robotics is a broader engineering field designing physical machines for tasks, often with pre-programmed or basic automation, not always requiring advanced AI (e.g., factory robotic arms). While EAI prioritizes cognitive adaptability, robotics centers on mechanical precision and task execution, with EAI being a subset that leverages embodiment for smarter, more flexible systems.Three Takeaways That Challenged My Thinking
- The humanoid form is overrated. “We don’t need a humanoid standing at the sink doing dishes,” Mohan argued, emphasizing function over sci-fi aesthetics. Zehan advocated for designs like “soft core with multiple limbs on wheels” that prioritize efficiency. This resonates with UK EAI research, where academics draw from nature—like the 1940s cybernetic tortoise that navigated without a human-like form. Paddington Robotics’ focus on general-purpose, non-humanoid systems and Wayve’s autonomous vehicles embody this ethos. The lesson? Let the task dictate the design, not Hollywood.
- The hardware ecosystem lacks standardisation. Mohan and Zehan highlighted a critical bottleneck: robotics hardware is a mess of fragmented components, with “firmware in Chinese and no standards.” Unlike the PC industry’s plug-and-play revolution, robotics innovators wrestle with incompatible parts, pushing them toward suboptimal humanoid platforms. The UK’s AI sector echoes this—53% of firms in a 2023 study flagged resource access as a barrier. Yet, the 2025 AI Opportunities Action Plan promises a 20x compute capacity boost and AI Growth Zones like Culham, which could drive standardization. For now, companies like Paddington Robotics must navigate this chaos strategically.
- Edge cases are the real battleground. Mohan nailed the data challenge: it’s not just volume but diversity. “Once 50 robots are in the field and one is in a bright sunny room not working right, how quickly can I fix that robot in an IP-controlled client environment?” This is where EAI lives or dies. UK researchers tackle this with frameworks like Deep Evolutionary Reinforcement Learning, helping robots adapt to unpredictable settings. Imagine’s work on diverse sensory data and Wayve’s real-world autonomous testing show how EAI must evolve. For founders, this means prioritising adaptability — your robot’s smarts need to keep up with the world.
A Note for Would-be EAI Founders
For founders diving into Embodied AI, the UK is a goldmine: tap into DSIT’s £1.3 billion AI investment or the AI Playbook to find high-impact use cases, from healthcare to infrastructure. Maintenance is your secret weapon—robots aren’t built for easy field repairs, so plan for it. Start small, solve real problems, and leverage the UK’s vibrant ecosystem to scale your vision.Shaping the Future of EAI
The UK’s EAI revolution, fueled by £1.3 billion in investment opportunity and innovators like Paddington Robotics and Imagine Health, is redefining robotics as adaptive, intelligent systems far beyond humanoid fantasies. Zehan and Mohan’s insights—ditching outdated designs, tackling hardware chaos, and mastering edge cases—point to a future where EAI solves real-world challenges with precision and flexibility. As the UK leads with initiatives like the Culham AI Growth Zone, the question remains: how will we balance UK leadership and innovation with ethical and practical hurdles to make EAI a cornerstone of tomorrow’s world?
Published April 24, 2025
HopalongPetrovski
I'm Spartacus!
Hi Bravo.
I have had a few run ins with the dumb mechanism too.
It has very limited intelligence and virtually zero understanding when it comes to an appreciation of subtlety or ironic humour.
It has either been designed that way intentionally as a hammer, or more likely in my view, what we are seeing is the limitations in programming inherent in current AI moderation.
Whoever is currently running this forum either does not have the will or the capacity to monitor it more effectively and perhaps more fairly.
But even where human moderation is used, such as over on the crapper, or in times past on Facebook etc, it is still rife with both ineptitude and obvious bias.
Beyond that, differing people will have differing sensitivities and flashpoint's in regard various topics due to their particular lived experience and so one's sense of humour, like one's appreciation of art, or what is and isn't, appropriate, is very subjective.
We tend to like people with similar predilections and it takes a degree of emotional maturity to maintain both tolerance and equanimity in the face of brute extremism.
Hang in there B.
We'll all be happier when BrainChip's success is better reflected in its share price.
Since my posts were deleted but Bravo’s weren’t, let me just say something in my defense… I just find it astonishing how someone can get so emotionally worked up over a simple illustration. It’s just an illustration… a joke… not meant to be taken seriously at all. Like I already said… some people here are acting like I personally slapped her... Come back down to reality. This is ridiculous… honestly. I can’t believe that this is a serious topic for some people…
You should think about it... if I punch you in the face as a joke, is it acceptable? No
Stop trying to defend yourself, you fked up, and the joke was inappropriate, you should go sit in the corner and think about it. Ask ur wife, friends, etc, or even chatGPT.
HopalongPetrovski
I'm Spartacus!
As I recall, you got pretty exercised when I posted Musk doing a nazi salute a month or so ago.
And seemed happy to be provoked when I posted a clip of an economist giving his opinion regarding the state of world affairs a few weeks ago.
We all have our hot buttons.
After realizing how some users here compare real people with a direct impact on human lives and actual interactions, to something as harmless as an illustration that was never meant to be malicious (which I mentioned before), I’ve decided to say goodbye.
Good luck to everyone else, and have fun. My fun is gone today.
Good luck to everyone else, and have fun. My fun is gone today.
HopalongPetrovski
I'm Spartacus!
So long, and Thanks for all the fish.
Fullmoonfever
Top 20
Though it's unlikely to hit this past qtr, but will see, it appears the DOD has outlayed about $77k to BRN as part of the recent contract awarded under COVID spending. Whether that's already physically paid previously and the record is retrospective, we won't know till the relevant qtrly I guess.
Jensen Huang says that autonomous driving needs machine learning:
12:00 >
But he gets the answer wrong:
ML needs massive compute = Blackwell!!!!
https://wccftech.com/nvidia-fully-e...ompletely-different-architecture-from-hopper/
I'm not sure how much of that power is needed for inference and ML, but even if it's only 10%, that's more than 100 Akidas.
12:00 >
But he gets the answer wrong:
ML needs massive compute = Blackwell!!!!
https://wccftech.com/nvidia-fully-e...ompletely-different-architecture-from-hopper/
NVIDIA’s Fully-Enabled Blackwell B200 GPUs Consume Up To 1200W, Completely Different Architecture From Hopper
Hassan Mujtaba•Mar 22, 2024 at 09:32am EDTI'm not sure how much of that power is needed for inference and ML, but even if it's only 10%, that's more than 100 Akidas.
Similar threads
