Over the past three months, I have attended A3 - Association for Advancing Automation 𝐇𝐮𝐦𝐚𝐧𝐨𝐢𝐝 𝐑𝐨𝐛𝐨𝐭 𝐅𝐨𝐫𝐮𝐦 in Memphis, IEEE RAS HUMANOIDS Conference in Nancy (France), and the Humanoids Summit in Mountain View. Here are my takeaways. 𝐇𝐲𝐩𝐞 𝐯𝐬. 𝐑𝐞𝐚𝐥𝐢𝐭𝐲 One would expect there to be a lot of cheerleading and hyping up humanoids at these conferences. That was not the case. While there were some presentations that leaned into the hype, the majority were very grounded in reality. The majority of attendees were fascinated by the tech, but openly talked about the issues that need to be solved if we are going to make these robots a reality. At this point, the more you are hyping up your humanoid, the more than likely you are farther behind on the tech and a chance of a deployment than those that are not - because they are too busy making their humanoid a reality. 𝐂𝐡𝐢𝐧𝐚 𝐈𝐬 𝐀𝐡𝐞𝐚𝐝, 𝐁𝐮𝐭... There is no denying that China is way ahead in humanoids compared to everyone else. However, that is mostly due to a massive push by the Chinese government with what appears to be an unlimited money appears. That is not going to last and if anything is showing the massive bet China is making here. Over these conferences, it became clear that China is trying to form one or two scenarios that give them the ultimate win in humanoids. The problem is there are potentially a dozen scenarios playing out and if any of those scenarios play out, China is going to lose and lose big. Unfortunately, it could take everyone down too. Not because others followed China down the wrong scenario path, but it will be such a massive implosion for the industry. See the impact of the Chinese Real Estate crisis on global financial markets. Time will tell. 𝐆𝐞𝐧𝐞𝐫𝐚𝐥 𝐏𝐮𝐫𝐩𝐨𝐬𝐞 𝐕𝐬. 𝐔𝐬𝐞 𝐂𝐚𝐬𝐞 𝐒𝐩𝐞𝐜𝐢𝐟𝐢𝐜 The early winners in humanoids are going to be the ones that nail great use cases for humanoids. Those right now trying to build a general purpose robot are just going to eventually run out of money before achieving that. The tech runway is simply too long for a general purpose robot than any VC is willing to fund. If humanoids are going to be successful in the near future, they need to be "employed" doing very specific tasks that no one wants to do and employers can't find people to do. This is still Robotics Business 101 right now and humanoids are not going to escape that reality. The clock is ticking and the money will eventually run out from both VCs and Governments. The race is on and there will be winners and a lot of losers. We should still see where we are going with humanoids, but we need to do it in reality and being honest with ourselves and the rest of the world.
Derek Black
1w
Aaron, why do we need humanoid architectures at all if we’re going to deploy them only for specific tasks and tailor them to use cases? This seems contradictory to the fundamental argument for humanoids versus traditional approaches to automation. Wouldn’t it make more sense to configure far simpler tools and software to do specific jobs?
Brian Wilson
1w
I'm curious to know what were the tasks that they said that they were so good at that would replace current automation solutuons in industries like manufacturing, logistics, healthcare, etc. Looking forward to the next post.
Aaron, I’d love for you to expand your thoughts on China. Also, it seems the case for humanoids vs other types of robotics are their general purpose use. Great post!
Aaron Prather I am surprised how little people talk about military use of humanoid robots. The form factor is great to drop from a plane and to steal enemy equipment or sabotage behind enemy lines and the Ukraine/Russia conflict has shown that modern well equipped armies like Russia lack enough soldiers to refill the trenches on the front line with training being a key bottleneck where humanoid robots are built and programmed. Armies would place 10,000 or 100,000 unit orders greatly reducing costs and DARPA was key in getting humanoid robots to to top of mind in the robotics industry. What am I missing? Is this quietly happening behind NDAs? It seems so obvious and ethics has not stopped other military weapons development.
Maybe you can explain me the benefits of humanoid robots? Is walking on two legs is an advantage? I don’t think so. Do we really need to create something that looks like a human and maybe later on thinks and feels similar as a human? I’m sure we will not stop this development and we will enhance our efforts to the area of genetic improvement and will combine this with the robotics. The result will be a new generation of self replicating living beings.
Wendy Rusk
1w
Aaron, this is a great article and appreciate the candor here. We support a lot of the robotics space with our technicians and engineers and it’s been like having a front row seat to watch these companies “train” their technology for various use cases. I believe a couple of the companies will rise to the top of the pile while others can’t scale fast enough or prove out the ROI. While I strongly believe in the use cases for humanoids they will need to overcome two major hurdles: getting out of the cells and working along side humans and other technology as well as get faster and the tasks they perform.
Thank you for sharing your insights, Aaron Prather! Helpful for those of us who were able to attend, and thos of us who were not!
Thank you Aaron Prather for the overview and state of the industry. Humanoid form factors are logical (most likely to fit into the human world), however the infinite capacity of the human to adapt is challenging to replicate. Every robotic technology will eventually find its "natural" application. Humanoid form factors will find their application, it may not be as generally applicable as we are currently thinking.
Calvin Holt
1w
China had a big push on controlling its production of humans and look how that turned out.
Director, Robotics & Autonomous Systems Program at ASTM International
1wAs one Humanoid C-Suite member stated at one of the conferences -