Home robot should cost $ 2999

Why do I say that? Because I'm on the path to creating such a robot. Because all the robots that are now are either toys, or scientific developments outside the walls of universities, or military developments, or the fruits of startups. Or vacuum cleaners.
There are several robots that may well claim to be a home robot. For example, PR2 from Willow Garage. Here he is:



Do you know how much it costs?
')

$ 400,000

And you know what he can do? Very little…

And this is the main reason, which still does not allow each of us to have a personal robot at home, and be limited only to a vacuum cleaner ... More precisely, reasons 2:
1. Buy a robot expensive
2. The robot does little for such money.
It turns out a vicious circle - in order to write software for a robot, you need to purchase very expensive hardware. And to buy iron, which is not able to do anything, is expensive. I will break this circle. Loud statement? Perhaps.
Here are the basic prerequisites that allow me to think that in the next couple of years, anyone can collect it himself or buy a robot that can do something really useful for $ 2,999:

• The appearance of Microsoft Kinect worth $ 200 and the SDK to it (seriously, without it, I would not even start)
• The emergence of a large number of developments in the field of applications for robots, pattern recognition, navigation, etc. For example, ROS, OpenCV, etc.
• The emergence of microcontrollers and sensors to them with a low threshold of entry into the development and programming (Arduino)
• The presence of a large number of standardized elements, such as servo drives, gearmotors, stepper motors, video cameras, batteries and chargers for them, etc.
• Increasingly widespread use of 3D printers
• The appearance of cheap and quite powerful laptops in a niche up to $ 500
• Wonderful open-source development inMoov, which allows you to create a mechanical arm in almost all similar human (this is the second thing without which it was not worth starting) - the video below
• Speech Recognition and Synthesis Services and APIs
• The cost of this whole economy fits into $ 3000!

And here I want to make a small remark. I am ready to create an open platform robot for $ 2999. But you have to understand that I cannot write all the software for it. The goal of the project is to enable the creation of a unified cheap open-source robot in order to develop software for it for everyone . And there at least sell it, even distribute it in vain.
The basic principles underlying the creation of a robot platform:

• All parts of the robot are unified and cheap. They are mass-produced and sold in stores.
• All plastic parts can be printed on a 3D printer, which does not exclude their production by other methods. I am ready to set up their industrial production with a sufficient number of pre-orders or upon receipt of the initial investment.
• Assembly of the robot and loading of software should not be more difficult than assembly of any IKEA cabinet
• The robot has a size comparable to that of a human and the hands are similar to human
• At a basic level, you must comply with the laws of robotics and try to protect yourself from malicious programming of robots (yet mass + hands can be a serious weapon or tool)
• The robot is controlled by voice, as well as feedback. In addition, the robot can display information on a laptop display.
• Robot programming = robot control. Those. Programming is carried out using voice commands. Each command at a low level is implemented in the form of program code.
• Also, the training of the robot can be done with the help of "gloves" - wearing such gloves, a person can perform movements that will be converted into a robot program. Gloves, however, will have to buy separately - about $ 500 per pair. (Yes, they are much cheaper than the proposed solutions and remove information from the whole arm starting from the shoulder + fingers)

Nevertheless, I perfectly understand some doubtfulness of the acquisition of such iron. Yet big money - 3-4 salaries for most compatriots. Therefore, the first thing I propose to do is a farm of robots .
A farm is a room - one or several, in which there are several robots. In addition, the room may be typical objects - tables, chairs, dishes, etc. Ideally, some typical “as in an apartment” environment should be created. It will be possible to connect to the robot via the Internet and upload the program to it. Then execute it. Feedback will be provided by video transmission from the board of the robot and from the webcam. Also, it will be possible to communicate with the robot by voice, for example, via Skype.

Creating a farm will allow thousands of developers to create and debug in real-world software for the robot. I would like it to be open. Then anyone can assemble his robot and upload this software onto him. And then it will be possible to talk about small-scale production. Then it will be possible to make the App Store for robots ...
Generally speaking, a lot of useful functionality can be written with only Kinect connected to a computer.

What does the robot consist of:

• Microsoft Kinect
• A laptop
• Chassis with 2 driving wheels and one swivel. 2 gear motor for wheels
• Plastic case, which provides for the installation of specialized extensions in separate compartments
• Battery + Charger
• 2 InMoov hands (plastic, wires, servos, arduino)
• "Nervous system" based on Arduino, Motor shield and a set of sensors

However, why do I appeal to you and ask for help? What do I already have? I tell and show:

• Created a prototype of a robot without hands. The nervous system allows him not to stumble on obstacles and navigate in space. The software is still being developed, and in this I hope for your help. Now in the development of "recognition" of persons, gesture recognition, voice control.
• An agreement was reached on the production of Arduino boards.
• An agreement has been reached on the production of plastic parts for the manufacture of robot arms. Now, with single runs, making plastic for one hand will cost about 50,000 rubles. To reduce the cost of production you need to increase the mass.
• Here's a render of the future robot (yes, it is a little different from the prototype in the video):

So, how can you help,% usernamre%?



And also, if you are a robot, programmer, designer, 3d-modeler, PR specialist, journalist, philanthropist, investor, manufacturer, and many more.

If you send your old laptop or the Kinect dusting in the corner, the Arduino set you played with and tired of it, if there are servos or gearmotors, some unnecessary sensors, wheels from lawn mowers, unnecessary children's electric cars or cases from household appliances - you really really great help. And we will not forget you.

What will you be for it?

• Priority access to the roboofarm
• Putting your logo on the robot
• Mention you in the list of people who made a donation + link to your site
• Thank you very much
• Will that tell grandchildren

A little bit about yourself. My name is Alexander. I am 34 years old, I am a little idealist, but to a much greater degree pragmatist and bore. I have quite a wide experience in IT. I was a programmer, tester, team leader, project manager. Now I am an entrepreneur, but still working in the field of information technology.

I came close to implementing the most costly parts of the project. I want to assess the relevance, clarity of the idea at this level. And yes, I want your help, money and PR. Because so the project will develop many times faster.

Summary: The main thing in the robot is not hardware, but software. In terms of technology, all the necessary frontiers have already been achieved. Creating a relatively inexpensive robot is simple. The main difficulty is to involve programmers in writing programs to control the behavior of such a robot. And for this there is a solution - to create together an open platform for the robot, on the basis of which everyone can write software for it. I invite you to participate in the creation of such a platform and create a farm of robots. Any programmer can use the robots on the farm.

There is a Google group https://groups.google.com/forum/?hl=en&fromgroups#!forum/home-robot
I don’t give the rest of the contacts, as this is not welcome ...

UPD. Thanks to everyone who responded and agreed to help in word and deed. I can now be somewhat sketchy in the correspondence, since I have a lot of communications, but I remember all of you and will answer everyone in more detail. Now I am engaged in the creation of working groups in different directions. With the advent of the site, all information will be available on it and you will be able to choose the direction of development to your liking. By Monday, I will give all the answers to the letters, if you have not already done so.

If you want to prepare yet with tz. programming and you are interested in the interaction of the machine with a person and the world around you, I suggest borrowing from friends or buying Kinect and connecting it to a PC, this will help you a lot in the future.