The biggest problem is they are using RC style servos. This is why all the arm motions are jerky and lack precision.

What this means: No force feedback on positioning. Jerky motion due to actuator steps. Limited precision. Terrible slow motion control performance. Fast movements will look better. Inability to solve problems with software.

Basically think of all the joints like those cheap thermal camera screens that are at 320 x 240 resolution. For rough work or finding some sort of oddity your good, but anything precise you need to spend real money. It's a pretty cool toy, but don't go buying this expecting that much.

Looks cool.

Can you be up front about its real capabilities. Are these videos of it eg tidying up real or just staged / cherry picked?

If you let it loose in a messy room or stirring something in a kitchen, what are the genuine results you would expect in the wild?

We’re all used to seeing things like Atlas back flipping but then can’t do anything in a real environment. Just curious how this fits in and what the success / fail rate is.

Where are you guys in SF? Would love to come see it. Mostly interested in seeing the speed of action (what multiple is the video replay at, etc.) and looking at center-of-mass etc. issues. I have a pretty straightforward use-case in a home. The Dreame we have in the living room is pretty good at avoiding most obstacles, but I usually have an LLM-based agent review the image automatically and either send out the vacuum or suggest moving things that would foul it up. Works great. Obviously substituting with a humanoid capable of reaching ground height things and picking up children's toys etc. would be great.

I'd want to bolt it with a mount to the wall when not in use because it needs to be climb-resistant for children during the daytime when I don't need it functional but there is a solid 8 hour 2200->0700 window when no one will be in the living room and speed is not an issue.

I was talking to my wife about this device and actually the wheeled form factor and (probably) no control software for balance is a feature for our use-case because when the Unitree G1 or whatever falls over the result is catastrophic. The balance control software attempts pure insanity and will likely kill a child nearby.

honored to welcome another nori* business, alongside

- nori.ai (ai health coach)

- heynori.com (ai family hub)

- nori inc (soil based carbon sequestration)

- noriagentic.com (that's us, ai cloud infrastructure)

and my personal favorite

- nori.co (makers of 'The Nori Cloud' a clothes steamer, as seen on Oprah's Favorite Things)

Website absolutely kills my S21U/Edge phone to view. Shame, this looks super cool. I'll check it out on my laptop later when I have a moment. Price seems enticing too.

My question would be how riffable is it? I don't have 10-20k to drop on a robot. I work for co-ops so while I make good money(90k~ pretty good!) its less than I could make else where but the trade off is I have more time to build random stuff. 2k is the price point for me where I might buy a big thing every so many years, teaching this to play board games, making new hands for it, teaching it to clean etc would be fun and at a price point that works for me. I would probably want a suction attachment for the fingers so it could handle cards and other flat tokens but that seems simple enough to build given the 3d models for printing are included.

I could see myself printing all sorts of costumes for it and using it as an ad hoc characters for tabletop games or as a spare bandmate if I ever finish my synth.

"We make money by selling the hardware for $1,688, with optional paid software on top."

The software would have to be quite good and also affordable for me to consider it, as all subscriptions are liabilities. I might one day want my robot to go with me to a show or to do something meaningful. Whenever that happens I will NOT be held hostage by software and will either source my own software or replace parts of the hardware till something works before I keep throwing money at a bad service. I would prefer to just have a good vendor supported software that I don't have to think about but I have trust issues. Privacy would be a huge concern, I can't have this in my garage lab with all my stuff being sent back to your servers for computer vision tasks etc.

I'm confused, is this a research device, hobby/tinker item, or consumer product?

I don't mean to be negative, but making a sustainable business on a hardware race to the bottom sounds challenging.

I'd be curious to to know from others here, what in-home tasks do people think they'll really be able to automate with any of the various bimanual or humanoid robot platforms that seem to be going around?

RPi5 seems a bit underpowered for all the sensors & control and offloading to a computer through wifi seems problematic. A Jetson Orin Nano would seem to be a better choice for a robot. Unfortunately Nvidia recently substantially increased the prices for their jetson devkits due to component shortages (DRAM,storage, etc), and what was $250 now costs $499 :(

This is cool! As someone who's built based off of old Roomba platforms and other models, I feel like this looks pretty fun and a good price point. What do you think the next version will target? Will you stay educational and low price point? Targeting some other usecases?

None of the videos on website say how much are they speed up (or not), and is robot teleoperated or running autonomously.