11/27/2013 | 1 MINUTE READ

Precision Robotic Bin-Picking

Facebook Share Icon LinkedIn Share Icon Twitter Share Icon Share by EMail icon Print Icon



Facebook Share Icon LinkedIn Share Icon Twitter Share Icon Share by EMail icon Print Icon

Related Suppliers

The problem: Removing randomly placed parts from a bin requires a complex integration of image recognition software, robots, end-of-arm tooling, robot accessories, and a sophisticated handling strategy. And once the piece is selected, the object must be precisely positioned on the relevant production or assembly line. This capability, called random bin picking, has been difficult to reliably  execute in real-world environments. Reliability has been hard to achieve outside of a laboratory.

The solution: Liebherr (liebherr.us) developed a random bin-picking system that integrates computers, software and strategy to make automated random bin-picking more efficient and effective.

The Liebherr system uses a 3D laser scanner to recognize parts with a high level of precision. Software identifies and selects information about items and constraints in the bin provided. Reliable automation components by SCHUNK Inc. (schunk.com), including the rotary unit PR90, swivel unit PW90, and gripper PGN plus,  grip and move the items to the desired destinations. One problem with bin picking is that parts can be located close to the bin walls or otherwise difficult to get to; the Schunk rotary unit PR90 provides collision-free access and removal of items from a container, even in difficult situations. 

The bin-picking solutions are available for automotive and commercial vehicle components weighing between 1 kg (2.2 lb) and 50 kg (110.2 lb).

An in-process inspection method is accurate to 0.1 mm.  If a recognition failure occurs, the system is delayed for only a short time thanks to an automatic correction strategy built into the software. Process reliability is improved by advances in image recognition technology and computing power in standard PCs. 

The proof: Liebherr conducted reference projects with the Fraunhofer Institute (fraunhofer.de) to test complete systems for the automotive industry. One project included the channeling of non-rotationally symmetrical components, which arrive randomly in a bin and must be placed onto production lines and into hardening ovens. An important breakthrough was the increased capability of the object recognition system, which can discern black, brown or rusty parts even with the presence of outside light.

This industrial scale automated bin-picking system increases output through higher productivity; allows workers time for other requirements; and increases reproducibility/consistent quality, according to Liebherr.