The Industrial Bin Picking Revolution

In my last post, I discussed the differences between a Pick and Place vs. an Industrial Bin Picking application. How difficult it is to solve such a complex problem that requires going from an unknown part position to a known part position? The big question: Why has no one solved industrial bin picking and why has industrial bin picking only been a myth in the industry for so many years? 

The Right Approach to Bin Picking  

For the past 30 years, some of the smartest minds in the industry have tried to solve this problem using traditional methods. These robotic and machine vision experts would use custom end of arm tooling and would spend 3-4 months to solve only one part for a particular application. After all this time,  there was no guarantee that the application would even work. This would sometimes cost companies almost half a million dollars in the hopes of solving just one part. These barriers put industrial bin picking out of reach and labeled too risky for the average company. Only manufacturing giants such as Toyota, Coca-Cola, and Proctor & Gamble could even attempt such a feat.  

In the past 8 years, many machine vision and end effector companies have attempted to create a successful Industrial Bin Picking solution. However, neither machine vision or end effectors actually solves the problem of telling a robot to go from a random arrangement to an ordered arrangement. Industrial bin picking a motion control problem. This is where the power of Actin comes into play. 

Actin Simplifies Industrial Bin Picking  

Let's explore the infinite number of solutions needed for Industrial Bin Picking to be successful and how Actin solves the problem differently than anything else on the market.  

Let’s use this standard electrical conduit outlet as an example; we will use a simple suction cup to grasp it from the top. Imagine that this part is in the top left corner of our bin. If we tell a robot to pick it up, it's pretty easy right? If we  rotate that part 10 degrees in the same spot, it is another robot pose.  It would need to be rotated every 10 degrees for a full 360 degrees. The part would need to be shifted to the next quadrant of the bin, and then the next, and so on- which is far too many robot poses to program. 

Yet this is only one face of the part, you would need to program every 360 degree pick for every quadrant of your part, and for every face of your part! There are an infinite number of solutions and this is only for one part! How do you solve for infinity?  

The short answer is you don’t! This is why so many companies have tried and failed to solve this problem. The issue isn’t identifying the part, it is the 40,000 lines of code an expert engineer needs to write. Human error can also cause the lines of code to not work. 

Energid used the power of Actin to solve this problem in a completely different way. Instead of you, the user, trying to solve the problem, Actin enables you to describe the problem to the robot and let the robot figure out its own solutions.  

Instead of telling the robot how to move, Actin allows the user to describe the tool, environment, bin, and part to the robot. You tell the robot how you would like to pick up the part, where you would like to place the part, and Actin figures out the rest. 

Before Actin, it would take 3-4 months to program just one part, now you can program a part in 3-4 hours!  If you want repeatable “Lights out manufacturing”, take a look at the video below to see it in action! 

Energid thrives on impossible problems. There is nothing that is too complex for Actin to handle. If you would like to learn more about the power of Actin please click here. If you would like to talk to an engineer, click here.