The human system weighs 250lbs and can be placed anywhere. Let's ask what it takes to walk the factory robot in that direction. First let's have the work piece be moving, let's say on a conveyor belt. The old robotics way of thinking would be to introduce this variable into the programming of the bot/station, create simple sensors for either the work piece or conveyor itself to indicate to the programming loop where the part is with as little error as possible, and continue to keep accuracy while maintaining as much precision as possible using rigidity (which equals mass and space). Now the whole system is functionally 7 DOF, and you add in the error and failure modes of the 7th DOF (the conveyor system) and accumulate some error. Now just imagine instead of a conveyor the part is on a rolling table with random Z height, and so it the robot arm, and you can see this will fall apart, you can't fight this battle with deterministic programming, machining precision, and rigidity. Obviously if you extended this system to be a humanoid robot on a 3 legged ladder which would be 30+ DOF between the weld and the ground, it couldn't possibly work.

But back to the hungover human, why does this system work so well? The human has very good eyes and a very good internal IMU. They are looking at the end of the filler rod and the weld pool, and even though the information isn't that good coming through the scratched welding helmet, they can compensate for all that error and run an internal function that holds the torch and filler rod in the optimum position to do a good TIG weld while ignoring or automatically adjusting for tons of other variables. Now to address your original question, in our system 1. Are current cameras good enough to get an equivalent amount of information about the weld that the hungover welder has? Yes, in fact can get more information than a human can 2. Are IMUs as good as a hungover human has? Hard to really know, but seems like it, though if you need many IMUs attached to different limbs on a robot its probably not as good as humans yet 3. Is the power density of actuators and power storage good enough to approximate this 250lb system of a human on a ladder with some combination of DOF that reaches a sufficient range of motion to emulate the humans hands (whether the robot looks like a human or not?) - yeah, plus in this case the welder is plugged into the ground for the human anyway so that system is already attached to mains power

So given all this, seems like the limiter is just software, which is the bull case for this prospected robotics revolution

I personally am not bullish on 1:1 human hands either, but IMO the question shouldn't be $100k 2 ton Kuka arm vs biped with hands, it's overactuated robotics (build it from the floor with hard coded operations) vs underactuated (build it from the contact point of the work backwards with ML and sensors). We shall see which form factors prevail, but the type of robotics development posted here seems like the way forwards regardless, an ecosystem of small, power dense, reliable, accurate QDD actuators will lead to many general purpose robot applications. I recognize I am not using underactuated vs overactuated in their strict definition here but if you are familiar with robots I think you'll understand where I am coming from as far as a robot design ethos.

I will say though in designing robots of this type without necessarily being bound by trying to make a robot look like a human, I have often found myself accidentally recreating human arm DOF in a round trip way, it does just end up being well packaged beyond the "world designed for humans" talking point. Maybe hands will end up being a similar situation.

Check the results here - https://bringatrailer.com/ferrari/550-maranello/ Example EU market car, imported 2023 - https://bringatrailer.com/listing/1999-ferrari-550-maranello... JP market car, imported to Canada 2018 then US in 2024 - https://bringatrailer.com/listing/1999-ferrari-550-maranello...

Another interesting case with ships is the Trieste and several other Russian oligarch mega yachts being held in Italy. Italian law requires them to maintain the value of frozen assets, so they are spending millions per month to keep these yachts maintained.

The space between those 2 things is where you have to decide what you are really trying to accomplish. The program you use will have an impact on what your result looks like, you see this in the evolution of product design alongside the evolution of design software (boxy cars in the 80s, soap bars in the 90s, and the last few decades of cars with flowing designs with body line defining creases which modern A surface modelers seem to draw you towards). I find parts made in Blender with my workflow often look a lot more interesting and visually pleasing, using edge crease/bevel modifiers and sliding loops around vs. using fillets in CAD for instance, they both aim to soften an edge, but look far different in the end. If you are only ever going to 3D print parts and never CNC, you are already fast in Blender, and part strength vs mass doesn't matter much (especially to a degree where you don't care about FEA), Blender is plenty viable to make printed parts with.

You can footgun yourself easily with both programs, but I find Fusion to be worse for this, half because of the UI, but using tools like sketch projection for me has caused really diabolical issues in the timeline. The whole trick to CAD is being very careful with the design intention as you progress forwards, which is hard to learn coming from 3D modelers where that doesn't matter much and you can just shuffle around non destructive modifiers. This might just be due to my own experience difference in the programs though, I definitely remember going down some roads in Blender I never returned from on meshes when I was learning, normally by either applying subdivision modifiers, doing too many loop cuts, or using a tri/n-gon somewhere thinking it wouldn't be an issue or I would fix it later.

That said, I do think it would be nice for people to note in pull requests which files have AI gen code in the diff. It's still a good idea to look at LLM gen code vs human code with a bit different lens, the mistakes each make are often a bit different in flavor, and it would save time for me in a review to know which is which. Has anyone seen this at a larger org and is it of value to you as a reviewer? Maybe some tool sets can already do this automatically (I suppose all these companies report the % of code that is LLM generated must have one if they actually have these granular metrics?)

Many years ago I was servicing Maserati GranTurismos and Quattroportes of which some use a ZF 6 speed auto transmission. Since the same transmission is used in Land Rovers, I would buy parts from the Land Rover dealer which was nearby. One time I went there and they didn't have any fluid for the transmission for Land Rovers, but they did for Jaguar. The fluid was identical, but on a different shelf, and cost a lot more. The parts department said that Jaguar uses a 3rd party parts distribution contract in North America, but Land Rover does it in house, so every Jaguar part, of which many are identical to Land Rovers, costs more. They could not just bill out a Land Rover part internally to their own dealer to service a Jaguar either (they were a franchise that repaired both).

1. Such a device will require significant local compute, generating a lot of heat. It cannot be too close to the body, and require efficient cooling. In the cowboy hat, the processing can be placed above the head in the bucket of the hat, and the cooling dispersed in a large surface area around the brim

2. Such a device requires 360 degree camera vision, thus cannot be a backpack or vest type design (which also bring heat too close to the body). It also must be close to eye level (cannot be shoes).

3. Has to be able to be worn in any environment, with any style. A cowboy hat is great for sun protection, and in the rain.