AI-enabled Pick-it-Easy Robot

Developed in partnership with the Californian AI start-up, Covariant, Pick-it-Easy Robot features AI and intelligent sensors that enable rapid self-teaching to recognise different packaging and features. The robot learns how to handle new items rapidly through processes such as 3D perception, few-shot learning and real-time motion planning. Rather than learning how to master specific tasks, Pick-it-Easy Robot learns general skills that it can apply to new tasks, in much the same way that people do. The solution is deployed with minimal downtime and can be operational within days. It is also available in a Robotics-as-a-Service (RaaS) model, enabling clients to secure efficiency benefits immediately while spreading the investment cost over time.

Picking with 99% accuracy at Obeta

Pick-it-Easy Robot fulfils orders quickly and accurately for the German electrical wholesaler, Obeta. The company's distribution centre in Ludwigsfelde, on the outskirts of Berlin, operates 24 hours a day for five days a week to ship more than 5,000 orders a day on a next-day delivery service. With Obeta’s warehouse receiving hundreds of new products each week – from light bulbs and electrical sockets to smoke detectors and circuit breakers – the company needed a system that could easily handle new items without disruption. Obeta has been impressed with the Pick-it-Easy Robot's speed of learning when it comes to picking unfamiliar products. The robot works alongside Obeta's manual fulfilment stations, picking and placing products from storage containers into order boxes. It achieves speeds of 600 items per hour, depending on order flow, and works without breaks. Handling some 70% of pickable SKUs, the robot picks with an accuracy of 99%, despite having to learn new items each week. 

Robot handles poly-wrapped fashion for GXO Logistics

In another application GXO Logistics, the world’s largest pure-play contract logistics provider, is using a Pick-it-Easy Robot for automatic insertion of film-wrapped fashion items into pockets at its fulfilment centre in Tilburg, Netherlands. This is the first use of this technology for automated pocket induction in fashion logistics. The apparel and e-commerce sectors have traditionally held a number of challenges for logistics automation. Not only does the range of goods continuously change through the seasons with huge variety in product size, shape, texture and weight, but many items are packed in polybags. Due to the floppy and reflective nature of plastic film, robots find it difficult to determine the perfect grip point and handle items reliably. Yet in omnichannel and e-commerce fulfilment, robots are the perfect solution for loading goods into sorter pockets, since warehouse staff often find this work monotonous. 

The Pick-it-Easy Robot at the GXO facility uses AI-based object recognition to enable it to identify the optimal gripping point and speed for each product. After selecting an item from a tote, the robot drops it gently onto a glass chute, where multiple scanners gather item information from its barcodes. The item slides onto a conveyor, which transfers it to a chute that loads it into a pocket, ready for fully automatic sequencing by a pocket sorter system. During this transfer process, the item is allocated to the pocket, which is traceable due to its RFID chip. The solution also features a sophisticated retry function, allowing any items not successfully scanned to be picked up by the robot from the conveyor and dropped onto the glass chute again.