Refinement of the Vision of Robotarm.org

Over the last 2 months of robotarm.org’s existence, I have had a lot of conversations with people about building robotic arms and I consistently heard 3 pieces of feedback:

  • Robot arm components are too expensive to achieve the $1-2k cost point that we want to achieve.
  • The clumsy, industrial programming centric, user interfaces of robot arms prevents me from doing what I want with the arm
  • Manipulators are the hardest part of building a robotic system

In order for RobotArm.org to meet it’s vision it needs to focus on these three items over time but since 3 is probably too many to get started with, I would like to see this community focus on the first two before tackling manipulators.

When it comes down cost of robot arm components, there are 3 expensive components (by my count, correct me in the comments if I am wrong) – the power supply, the joint – motor, gears, feedback, bearings (which is multiplied by the number of Degrees of Freedom that the arm has), and finally the control board/motor controller.  While the Power Supply is pretty much a commodity, the joints and controller are not and therefore enjoy rich vendor margins.  It seems that focusing on those two components as opportunities to reduce cost, will yield the greatest benefits.  While the joint is already the area I am personally spending all of my time, the controller board has not had a lot of energy spent on it to date, but I know @tlalexander is looking at it (yay!)

The second piece is a simpler problem in some ways… Because we are not building a “product” and instead are building tools for engineers, we can skip the interface for industrial engineers and build abstracted interfaces.  As can be seen in the scientifically derived pie chart below… we should focus on the green part rather than developers that specialize in robotics.

Screen Shot 2015-06-22 at 12.40.19 PM

As robotics meet the mainstream and developers want to take advantage of automation, the status quo of robotic interfaces creates a barrier for developers (it is worth noting that cost is also a barrier) to get started exploring robotics as a solution to problems.

I would love your feedback on these thoughts.  Just put them in the comments.

Leave a Reply