Industry Videos

Machinery & Metals

3D picking of transparent bags

AccuPick3D uses deep learning recognition to identify each bag and place it in the predefined location. This allows Solomon’s solution to pick transparent bags containing metal pieces that have been randomly placed in a bin using a pneumatic gripper with high accuracy.

SOLOMON Random Bin Picking extremely small objects

Solomon AccuPick 3D offered the customer a single software package that was able to recognize 50+ different parts in a working station. Using the power of its AI software, AccuPick 3D was able to recognize a 5mm sized washer, even when there were overlapping objects. Being able to recognize pieces of such small size has always been a challenging task for 3D vision solutions, but AccuPick 3D actively solves it by relying on a single vision software that allows it to full the needs of customers, helping them to lower operational costs and maximize pro ts.

Solomon joins as member of MVtec Image Acquisition Partner

With the objective of shaping the future of material handling together, SOLOMON is growing its network of partners around the world.

Yaskawa Motoman Bin Picking (with Collision Avoidance)

A Yaskawa robot picks metal objects and distinguishes front vs. back sides using Solomon´s AccuPick 3D. Motion planning is included to prevent the robot and gripper from colliding into the bin.

Deep Learning Inspection Using Only Golden Sample (GS)

Images of defect-free PCBs (GS) are used as a training set and input into Solomon´s AI inspection software. Unlike in conventional automated optical inspection (AOI), it´s not necessary to write a code or train it with images containing individual defects.

Bin Picking of M3 Screws with a Mecademic Robot

As displayed at the 2018 Stuttgart Vision Show, this 3D vision system is trained to recognize object´s orientations using deep learning technology which allows it to effectively pick M3 screws from the bin.

Bin Picking of Complex Shaped Objects

The robot is trained to pick complex shaped objects via deep learning combined with 3D vision. No CAD file is required to match the object´s 3D pose. The training time is significantly reduced compared to the conventional template matching methods.

Bin Picking of Two Different Objects in a Bin

The robot is trained to pick and sort two different items using our 3D scanner and software, along with a 2D camera for stacking alignment. For this demonstration, a Fanuc robot with a payload of 7kg was used.

Bin Picking for Molding Machine

A UR5 robot is tasked with picking round shaped metal pieces that have been randomly placed in a bin, to later insert them into a molding machine with synchronization of the control systems.

Bin Picking of Black Screws

A UR5 robot is tasked with picking M4 sized black screws that have been randomly placed in a bin.