Order picking robot

An order picking robot takes over repetitive gripping tasks and also works where light and visibility are poor. In the Fizyr and IDS example, workplace safety improves and the low error rate reduces logistics costs. Because the software detects damaged parcels, they do not end up on the sorter. Users decide for themselves whether to hand over individual tasks or automate entire processes such as unloading lorries.

The robot combines a 2D image with a 3D point cloud and lets software classify the objects. In the solution example, a uEye CP camera supplies the 2D image and an Ensenso camera the spatial coordinates. Fizyr's algorithm recognises boxes, bags, envelopes, tubes or cylinders it has not seen before, analyses their surfaces and proposes the best gripping position.

Yes, with precise image processing an order picking robot can load and unload cargo areas. The load space is the hard part, because light is often poor and parcels overlap, are tightly stacked, glossy or transparent. Fizyr's software analyses images from IDS cameras for this and calculates the gripping position. A wide camera field of view and high camera speed give the robot short reaction times.

Yes, in the solution example depalletising is one of its tasks alongside picking, palletising and handling a wide range of parcels. This works because the software recalculates every gripping position from camera images instead of moving to fixed positions. The robot needs many degrees of freedom for this, six in the example, and performs dozens of grasp positions per second. Integrators choose the hardware for their cell freely.