Beyond science!

Yes I realize batteries is the field (along with fusion, hydrogen power cars etc) where false promises breakthroughs are made like every second week.
So I'm not sure if this was the one, I was thinking of. I thought there was an ozzie link, so this is what I came up with a quick google, I'll have a look later to see if this was the one.

EDIT: Yes I think this was the one I read about some months ago, I guess we will see if it pans out, cause there has been a lot of unfulfilled promises in this field, one difference I see is they want something out in < a year unlike the 5-10 years that normally get banded around. Also it has the university of brisbane involved, which I assume dont want to sully their name based on a cash grab.

 

seems like it was only a few years ago and they had those loud engines and were attached by wires to the ceiling.
Any idea how long they can keep this up for before the batteries run out? i.e. is it closer to 10 minutes or 10 hours? I'm guessing the former, but have no idea

 

Atlas runs for an hour off an 3.7Kwh battery. It's not a product though. You'd probably ring more out of something that was. Spot does 90 minutes.

 

When the batteries become more efficient these bots would be ideal for space/planet exploration.

 

Spot can return to its base station to charge, so you could potentially have a big rover as home for a few limbed robots to go exploring things in more detail. No one's proposing that for an actual mission yet. JPL are a Boston Dynamics customer but not sure what they're doing with it.

Tesla have obviously announced their 'Bot. That'll inevitably end up in SpaceX's efforts if it works. They have a lead on battery efficiency. Using more efficient mechanical actuators is viable as well, in a way it wasn't when Boston Dynamics started.

Autopilot failures obscure the fact that Telsa's neural net world perception/machine learning/navigation stuff is pretty amazing. Humanoid robots could become useful beyond tech demos if you can leverage that.

A humanoid robot much more limited physically than Atlas could do useful jobs with a good enough perception of it's environment.

 

Cue musk telling everyone it will be cheaper than a roomba, do all your housework and when your not using it it will go out and freelance for work earning you $20k a year and it will be on sale in 2023

 

He said it's could eventually go to the shops for you, which is even more far flung than housework.

Like most of the soundbytes Musk generates, it's not really relevent. The bottom line is that Tesla may end up with the first actually useful for something hominid robot, that actually ends up as a product and can be itterated from there.

 

Is it beyond science when the science is done and the engineering starts?

MIT/CFS have made a record breaking high temperature superconducting magnet. It puts small tokamaks with net energy production within reach (existing tokamak designs use lower temp superconducting magnets, which forces a massive size if they're ever to achieve net energy production).

They have a very serious chance of achieving net energy in the next 5yrs.

https://cfs.energy/news-and-media/cfs-commercial-fusion-power-with-hts-magnet

 

It's a huge step for them. They're not the only fusion group/company pursuing magnets using high temperature super conductors. Hopefully some of the others will also have prototypes working soon as well.

It's encouraging stuff.

 

Commonwealth Fusion have secured $1.8bn in their latest funding round. That should mean their '2025' SPARK test reactor, using the HTS magnets above, is pretty much guaranteed to be built. Obviously no guarantee it'll be on time or work...

https://www.cfs.energy/news-and-media/commonwealth-fusion-systems-closes-1-8-billion-series-b-round

 

I watched some of that guy's videos in the past and they were not that good. I probably should not write him off yet but I sort of have.