Dream projects?

[meltwater]

Just wondering what projects people would like to do if they had the skills, time and money to do them?

Top of my list would be a Bi-ped robot, simply because!
- BRL's robot (not RPi based but excellent all the same).
Then combine with some of this stuff:
http://sploid.gizmodo.com/even-computer ... @caseychan

Then link to RoboEarth:
http://www.bbc.co.uk/news/technology-25727110

Love the idea of RoboEarth...wonder how if it could be adapted to something suitable for open source development, like git-hub but more fluid and interconnected (you design code to do something, that is available for other software to pick up and use just by querying the database).

[jamesh]

I've got an idea for a project involving me, Kylie Minogue and a large bucket of baby oil. I have the skills and the time, just not the Kylie.

That Siggraph paper on bipedal walking is really good btw.

[morphy_richards]

I would love to finish the hardware for the skutter, so it has got the physical potential to sense and move around in it's environment.
I'd then like to build a programming project based on Python and mysql that would let it take on objectives (like missions) and then record actions that it can take and the order that actions should be carried out in order to achieve levels of success.
Python would be used to make sense of the mission objectives and the mysql database used to record all and any possible actions the skutter is capable of performing.
Actions could then be grouped and associated with objectives and then more simple objectives could then be grouped to allow more and more complex objectives to be accomplished.

During "sleep" (battery charge time) the skutter could run automatic processes to allow database normalisation and perhaps run simulations to allow it to populate another database of potential actions to be tried during "awake" time.

All I need to do is win the lottery so I can concentrate on this and stop having to do this annoying "work" thing.

[meltwater]

I would imagine if everyone had their own Kylie [or indeed Robbie - alternative dreams are available], the modern world would probably collapse on itself (and baby-oil prices would be through the roof).

That Siggraph paper on bipedal walking is really good btw.

Do you have a link for that, couldn't find anything from the video when I looked.
EDIT: http://grail.cs.washington.edu/projects/loco/

The research groups have all the fun, http://www.gizmag.com usually gives me a daily dose of "wow, wish I could work on that". As an engineer though, just want to take all of it and put it into something awesome.

Yep...the "work" thing, gets in the way of the raw potential.

RoboEarth for skutter would be neat.

[AndyD]

I've got an idea for a project involving me, Kylie Minogue and a large bucket of baby oil. I have the skills and the time, just not the Kylie.

My parents house (where I grew up) is about 2.5 Kms from Pin Oak Court, better known as Ramsay Street where Neighbours is filmed. Now if I can just get that time machine project off the ground ...

[rant]We would never call a cul-de-sac a street in Melbourne, most of them are courts[/rant]

[meltwater]

I'm rather taken with this RS4 robot at the moment:
http://letsmakerobots.com/node/38610

The self balancing aspect of it is really good, certainly more interesting than a little buggy bot. Reminds me of the telepresence bots which we are all supposed to have by now.

[Ravenous]

During "sleep" (battery charge time) the skutter could...

Actually, for me I would just be pleased to develop something capable of plugging itself into a wall socket (any standard wall socket, not a specially marked "nest") when necessary.

As for the really ambitious stuff like washing the dishes, I don't think I'll ever get an automated solution. (Also bear in mind I would be too lazy to even load a dishwasher!)

[meltwater]

During "sleep" (battery charge time) the skutter could...

Actually, for me I would just be pleased to develop something capable of plugging itself into a wall socket (any standard wall socket, not a specially marked "nest") when necessary.

As for the really ambitious stuff like washing the dishes, I don't think I'll ever get an automated solution. (Also bear in mind I would be too lazy to even load a dishwasher!)

A UK wall socket would be quite a task, since they are designed to take significant force to insert and remove plugs from (with safety covers on the live pins which only open when the ground pin is inserted). Although would be fun to make a robot that could insert a USB plug the correct way around without trying three times.

Disposable plates is probably the easier (and possibly the cheaper options).

[meltwater]

The following caught my eye today:

Not sure if they are fitted out for light-saber practise yet though.

I didn't know they'd had similar ones buzzing around inside the ISS since 2006, what an excellent idea. Couple that with Robonaut and the 3D printing experiments, plans for capturing small asteroids etc.

Makes for an interesting thought, send a robot and 3D printer to remote place in the solar system, then task it with building various robots to explore and perhaps even prepare the place. Keep collecting materials, and building more things. Sounds plausible that we could have general purpose robot-manned ships working in various places in space (although what advantage does a robot have over a typical purpose built probe). What useful things could be created in these locations, which can then be useful?
But how awesome (aside from our natural nature to go to places and mess it up with our stuff)!

A common issue which keeps cropping up is that we are very poorly designed for space, particularly long periods and specifically when it comes to accepting the large amounts of radiation out-there (at least from my low understanding). However, it is something we will have to overcome (one way or another) if we are going to go anywhere.

The fact that travel is likely to take a very long time IS an issue, but that probably isn't that huge (given enough money it is feasible to build a self-sufficient environment), keeping people alive that long sure is. I wonder if this is something which will come from advanced materials or medically we adapt our bodies to, or even...getting far fetched here and into moral muddy water...are specifically bred for genetically...

[Ravenous]

I like the look of theose Spheres (Synchronized Position Hold, Engage, Reorient, Experimental Satellites). All I need is a zero gravity living room to try one Maybe I should look into helium prices...

Speaking of satellites, has anyone seen the reaction wheel robot? A cube that can jump up on its corners and "walk" around the desktop:
http://www.idsc.ethz.ch/Research_DAndrea/Cubli

[meltwater]

I like the look of theose Spheres (Synchronized Position Hold, Engage, Reorient, Experimental Satellites). All I need is a zero gravity living room to try one Maybe I should look into helium prices...

You say that...rather reminds me of a Dr Who christmas special with Air Fish.
http://psipunk.com/airpenguin-%E2%80%93 ... ing-robot/

Speaking of satellites, has anyone seen the reaction wheel robot? A cube that can jump up on its corners and "walk" around the desktop:
http://www.idsc.ethz.ch/Research_DAndrea/Cubli

Yes, I want some of those too. Will have to look at the videos later.

EDIT: Seen the videos...I want some cubes!

[meltwater]

[I will say now, I only have a passing interest in this, so I admit there are huge holes in my understanding and knowledge since I don't often have chance to read up on it all, but I do find it very interesting.]

I have space on the mind at the moment, particularly since I read about the concept of having space gas stations. The initial premise is to just to have a location where missions can drop off their "contingency" fuel, which can then be picked up and used by other missions.

This is rather than the current situation where each mission takes up their own (at huge additional cost) and is typically jettisoned prior to re-entering orbit. The "gas" station would be a gravity neutral location, avoiding the need to expend lots of energy keeping it there. Wonder if it would make sense to use it as a collection point for space junk too, collected from orbit and sent there to be processed, even as general mass I assume it could be used as a propellent. It would also provide a suitable place to process the asteroids they intend to capture.

Again, I think it will be far more realistic for robots to be used in this location to perform the processing. It could be imagined that multiple stations could be placed in the various gravity neutral points in the solar system, potentially making a series of stepping stones rather than one single long trip (or at the very least locations where things can be kept for future trips).

[Ravenous]

Great idea in theory. The ISS gets a visit every 3 months (I guess that's still the crew rotation period) but at the moment I don't know if there's a way to transfer fuel. And they only carry enough fuel for their own needs - and need to keep the contingency for safe re-entry (imagine getting stranded).

The station needs regular boosts in its (quite low) orbit and this is sometimes done by visiting ships giving it a push. But I'd expect that's also planned in their fuel load in the first place...

Also stuff launched into different orbits for whatever reason (like polar orbits for whole-earth observation) are on ther own, because it would take huge amounts of fuel to route the used launcher to another orbit. So you could only use excess fuel from launchers put on nearly the same orbit as your station.

The ISS could have been used as a general fuel store but its orbit is never going to be ideal for every possible application (of course it's in a good orbit for its own supply, reachable from the usual launch sites). For example boosting geostationery satellites and lunar missions might ideally have maybe an equatorial orbit as a starting point.

Now for the dreaming - there is (at the moment) thought to be quite a lot of ice on the moon - hidden in permanent shadow at some places on the poles. That could be converted to H2/O2 by solar power and shipped back down to a station in low earth orbit. That would be cheaper in the long term. Until that ice runs out, and massively expensive to set up of course... But that's probably the plan if people ever decide to do space properly.

[meltwater]

I think the original idea was for earth to moon type missions, where additional fuel would be taken for landing/take off attempts. They drop off excess fuel on the way past (a separate tank sounds most sensible).

Seems to me that it needs to be driven by getting some kind of commercial benefit out of it, like the fuel production on the moon. Although if you could actually make space equipment on the moon or even in space that could be helpful, assuming you can gather the raw materials locally. I guess the idea of 3d printing in space is along these lines, particularly if they are able to reuse the material.

I expect that would be most cost effective if that was driven by using robots though (easier to drop off and easier to keep running).

UPDATE: One bonus of living/working in space would be when you are debugging someone elses code, they can't hear you scream.

[meltwater]

Anyone else come across this?


Structure sensor
http://structure.io/#

I saw a video demo as part of Geekbeats CES coverage, looked really good. Although $350 is a bit much for me. I guess the Kinect with a raspberry pi could do this, linked with an accelerometer...although I've not heard of anyone getting the kinect working yet.

[Ravenous]

Very interesting - at first I thought it was yet another stereo system of some sort, but looking at the developer page it says "Infrared structured light projector, Uniform infrared LEDs" and 640x480 resolution.
http://structure.io/developers

EDIT: on further googling it seems to be basically a bit like a kinect in the way it works. (Emits a certain pattern of infra red, interprets what comes back.) Maybe the difference is they've made it more of a stand alone device?

[meltwater]

Yes from their descriptions and images looks like an IR grid or pattern, and some fast processing to make sense of it.
EDIT: Unlike kinect it appears to track movement information, so as you move the sensor it can build up a full object.

I just love the idea that the family photo could be a virtual 3d model of the kids.

Was looking into making a setup like this some time ago:
http://www.engadget.com/2006/12/22/davi ... r-scanner/

[Ravenous]

I just love the idea that the family photo could be a virtual 3d model of the kids.

I wonder how long it'll be before we get 3D home video recording, based on either this sort of IR system (possibly difficult outdoors) or a two-lens stereo approach...

[meltwater]

I think it isn't too far away. I think the lightfield sensors have some depth capability (there is a new sensor for smartphones which uses this), the clever bit seems to be using accelerometer/magnetometer data to build up a single model (unless they just use data matching to stitch them). Although, even 3d image from a single viewpoint is useful.

The description of the images was that they contain pixels which have RGB etc. information and a depth value (much like the huge pin display they had at the olympics), even with just that you get a lot of useful info about the world.

EDIT: If I did have $350 going spare, I'd be getting the Oculus Rift dev kit...ready for David Braben's Elite release (http://www.pcgamer.com/uk/2014/02/25/el ... pressions/).

[Ravenous]

Thought I'd dredge up this old thread with another high-end robotics project that'll be impossible for most of us:

http://robohub.org/robots-m-blocks/

Basically they're small cubes with reaction wheels inside - so they can be made to spin/jump and assemble themselves into larger structures. (I've posted something similar in the past, but the video on this has a better example.)

Looks great fun.

[meltwater]

Thought I'd dredge up this old thread with another high-end robotics project that'll be impossible for most of us:

http://robohub.org/robots-m-blocks/

Basically they're small cubes with reaction wheels inside - so they can be made to spin/jump and assemble themselves into larger structures. (I've posted something similar in the past, but the video on this has a better example.)

Looks great fun.

Yes seen these little things, rather neat idea. A hundred or so of them would be great fun, although charging the batteries might be a chore.

Wonder what applications there are for something like that, I'm guessing a chair/table example isn't it, perhaps a mobile wall for riot control?

[morphy_richards]

I've got this half baked idea that I could get some astronomical binoculars and attach a pi with normal camera to one eye and pi noir to the other for visible and infra red astronomy.
The binoculars would sit on the roof of my shed and automatically seek interesting close and deep sky objects to photograph and then upload them somewhere.

I don't think that's out of the bounds of possibility but I want it to be able to somehow choose what to photograph. Some sort of intelligent system that can pick out humans current astronomy interests.

Mabye by monitoring Twitter?

[meltwater]

Would be rather good, particularly if you used image stacking to improve the results. Would be fun to have a go!
Would probably need some fairly accurate motor mounts I expect.