The point I was trying to make was that this is great, but at the end of the day my final asset is a little simulator running on my computer that is basically only interesting to other robotics geeks and doesn't actually "do" anything. Some my earliest projects as a programmer were things like photo organization tools, online forum software, ANSI art editors, things that solved a tangible problem and that I could share and others could use. This aspect to it was key in keeping me motivated since it wasn't really a "homework assignment" that solved some abstract problem. I feel like there's a gap here.

I assumed the final asset would be your education.

But there is a real gap here: You probably don't have any problems that a simple robot would fix. You can buy a Roomba if you want. The killer app for robots is the autonomous car, and that is coming. No one has figured out what you want a robot for in that huge gap. Heartland has a nifty $20K robot that will do light handling tasks. You still don't want one.

If you can think of a robot version of the little software projects, good on you.

Is the autonomous car really the robotics killer app? I don't doubt that the autonomous car will reshape more than one industry (real estate, insurance, and transportation spring to mind), but I think proper robots might allow us to shed so many assumptions about what is possible that it's actually currently difficult to imagine all the things they will end up being used for.

It will save around 20,000 lives every year in the US alone. And prevent many more serious long-term injuries. What other household robot will have such an impact on families?

Don't get me wrong, self-driving cars are the human advancement that I am most looking forward to.

I think they will both be extremely big changes in different ways. Autonomous cars will be one big impact that indirectly touches many industries. Robots are going to get into everything, and the world will look very different at the end of it.

Robots will do farming, cleaning, food preparation, fire rescues, surgery, manufacturing. Basically, whatever parts of the world software can't eat, robots probably can.

I think a plain old "getter" would be fun. I'm on a chair and I want something to get that thing on the out of reach coffee table and bring it to me. Like a roomba with a camera and an arm.

it is however a really good question - not just good on him, but inspiring to us all.

(late edit: not Heartland: Rethink Robotics. They renamed before Baxter launched.)

I can think of a few projects that would be useful and fun, i.e. laundry-sorting bot, sandwich-making-bot etc., but they'd all be expensive and difficult to do.

A cheap, robust robot hand/arm would go a long way towards making this kind of robot-hacking accessible, but the bigger obstacle is that the generic computer vision building blocks that you'd need don't really exist yet (OpenCV is a good start)

The problem is that many tasks that seem simple (when you take millions of years of visual system evolution for granted), for example recognizing a sock in a pile of laundry, turn out to be pretty difficult.

Some of the theoretical problems do seem to be quite difficult, but the great thing about implementation is that as long as it works "well enough," no one minds if you cheat. For example, recognizing a sock in a pile is difficult, but what if the robotic arm just tries to pick it up anyway? From there it can weigh it as well as better seeing the size & shape of it in order to determine whether it is a sock. The difficult theoretical problem suddenly gets replaced with acceptable behaviour for the laundry-sorting application.

I like your outlook :)

On top of this, if a robot is doing the work it doesn't have to be the most efficient. Do it in stages... sort by weight or rough shape. Then a second pass does the "is this a shirt or pants" computation. Then sort those piles. Finally come back to "pairing socks". Or don't. It's already saved me a bunch of effort! Let someone else figure out what to do with a pre-sorted pile and I'll just add the firmware or another stage robot. It feels very "hacker ethic" to me!

You might be surprised at the software components which do actually exist:

http://moveit.ros.org/

Check out the videos on that page. There are a lot of obstacles to inexpensive robotic manipulation; the main ones are power-weight ratio and power-cost ratio of the actuators.

I think your statement can be turned into its own solution...

I agree that there is a real gap there - so a great contribution would be to fix the gap!

Compare hardware hacking. Until the last few years, it was pretty darned hard to get into. There have always been kits - e.g. the basic stamp, but they felt toyish. Lately though through a combination of Arduino, the explosion of hackerspaces, globalization allowing cheap and easy board design, and murky social forces entering the DIY electronics world has become a real thing. Gaps are being filled in at an impressive rate now... People are using Arduino to fix stuff that would have been a major project 10 years ago! Yes Arduino still has a tendency to feel like a toy (but less of one than a basic stamp), and, there are lots of projects that fill the space between it and DIY pocket computer... look at RaspberryPI or Beagle for examples - you can do a really good control system with them, python and minimal hardware knowledge.

So I suggest to you or anyone who wants to fill the gap - try to invent the Arduino of robotics kits. On your side there are several nice factors: easy access to 3d printing, widespread knowledge of the electronics parts listed above. Halfway decent free modeling packages (although engineering and simulation software may need some love). The opportunity is ripe!

I highly recommend watching a talk given by Erik Nieves from Motoman at the most recent RosCon because he talks about the state of the robotics industry and why this "gap" exists in a very accessible way.

http://vimeo.com/67044628

Two reasons why I think the gap between getting started and being able to build something useful exists is that robot parts are still much more expensive than we expect them to be and the fundamentals of robot movement are much more difficult to learn than we expect them to be. Unfortunately, I think those two factors discourage a lot of engineers early on and is why the size of this community is so small.

I feel very fortunate that I have a Kuka YouBot to hack on. The company I'm trying to get started with two friends wrote a proposal that got accepted into the sponsored track of a competition Kuka is running. (http://www.kuka-labs.com/en/network/innovationaward/) After setting it up, it's clear why this is a 20,000€ robot. But before you can actually DO anything useful with it, you have to be able to understand kinematics and dynamics as well as how to get path planners to talk to controllers. And then the really powerful stuff becomes possible once you incorporate computer vision, which is also not an easy topic to understand. It not surprising that most of this work is being done at universities by grad students.

The state of ROS today feels a lot like what Linux felt like to me back in the late 90's. There are lots of issues with dependencies and programs not compiling correctly. Documentation is severely lacking. But there's lots of enthusiasm from very friendly people to help each other work to make things better.

What makes me worry is that many of the advances in ROS came from Willow Garage and Google is buying up many of the companies that spun out of Willow whereas a lot was made possible in computer vision due to the sensors made by PrimeSense, which was just acquired by Apple. I hope the Google vs Apple rivalry doesn't slow the pace of progress in open source robotics that was making the dawn of non-industrial robotics seem not so far away.