The world of intralogistics has been gripped by a wave of automation for some considerable time. Its origins date back to the early days of Industry 4.0. There is much talk of robots here, which is often used synonymously with automation. This isn’t quite right, however, because the main requirement of robotics is that the systems possess artificial intelligence. Automation is based on standardized processes, while robotics also incorporates variable operations – in other words, processes that are especially relevant when it comes to the storage and retrieval of heterogeneous product ranges. Yet there is still some way to go before the challenges of handling differently sized goods with various textures and picking them smoothly by robot are solved.
Nonetheless, the first steps have been successfully taken. For instance, dynamic storage systems already provide the best basis for full automation at a later stage. To date, it mostly seems the case that goods are automatically delivered by conveyor technology so that they can then be placed in the allocated trays by pickers. From here, they are transported to their designated storage space in the vertical or horizontal storage lift using control software. If the item is requested, the system conveys it to the warehouse worker at their central workstation in line with the ergonomic goods-to-person principle. Long walks are no longer necessary and the modern software ensures that the correct parts are always retrieved. Picking precision is increased to almost one hundred percent and simultaneously accelerated thanks to support techniques such as pick by light. The picker removes the item and prepares it for shipping or further processing.
With a system of this kind, the human work involved is consequently restricted to merely removing the goods and transferring them to the designated bin. All processes otherwise already run automatically. This is where robotics comes into play, because it would of course be very interesting to get a machine to also perform these final manual steps. The economic benefits of doing this are obvious: If orders are picked by robots, the storage and retrieval process can take place around the clock. Furthermore, the robots guarantee maximum precision and speed when picking items, which significantly cuts retrieval times even further. Robotics also offers a way out in relation to the shortage of skilled workers, which is currently a problem in many countries.
Robotics not yet sufficiently developed for widespread use in the area of storage logistics
However, intelligent robotics is currently only used in isolated cases in the area of intralogistics. Yet what is the reason for this? For one thing, the cost of the available models is still at a level that many logistics operators are reluctant to pay. In addition, the reliability of the systems often leaves a lot to be desired. This is primarily due to the problem of gripping the goods, which requires maximum precision, especially when a heterogeneous range of products is stored. At present, the models simply do not possess a sufficient level of artificial intelligence to be able to sort the different shapes and materials and give the gripper the right instructions in each individual case.
The industry is nonetheless working hard to bring the robots closer to the warehouse. In doing so, interest is currently focused on two approaches. The first involves robots that retrieve goods from conventional shelf storage systems and independently take them to the picking station. Here, Amazon has a method that has been tried and tested several thousand times over: Its Kiva system, which it purchased some years back, lifts the entire shelf and transports it to the workstations. In principle, the method is suitable for employing in many conventional shelf storage systems. The gripping problem also doesn’t arise, because the devices only move the shelves. However, the downside is that the shelves have to be adapted accordingly so that the little robots can move them around. The cost-effectiveness of the scalable system is thus cannibalized a bit too much.
Another approach is used elsewhere at companies such as Magazino, in which Siemens acquired a stake some time ago. Here, the delivery robot Toru independently maneuvers around conventional rows of shelves and directly retrieves the items that need to be picked. With this solution, too, the difficulty lies in being able to reliably pick the heterogeneous items. Even though the technology is quite advanced here, further development work is needed to enable it to pick books, stuffed teddies, screws, bags of chips, and footballs with the necessary precision. Compared to Amazon’s robots, however, no big adjustments have to be made to the existing shelves, apart from limiting the height.
Fully automatic approach
The disadvantage of both of the above-mentioned transport methods, however, is that manual work is still required at the picking table. That is why further development work has been carried out to enable the robots to pick the items independently with no human assistance. These include models such as Baxter, which looks like something from a science fiction movie and is able to manage items extremely nimbly. Baxter also already possesses the necessary intelligence to react to heterogeneous challenges. Another solution is currently in development in the form of the picking robot Kado, which is better at handling a wide range of items. That’s because Kado has been designed to quickly and reliably identify the ideal gripping points on items in load carriers using state-of-the-art 3D camera technology. This system is also said to work with items that have not yet been entered into the system or correctly sorted in the load carrier. The picking robot’s sensitive gripper will then be controlled via the identified gripping points. In conjunction with an automated storage system, this would be a solution that would function without any manual work.
However, it is not just the intralogistics experts who are working on advancing the technology. Amazon, for example, has (not completely unselfishly) launched the Amazon Picking Challenge. Participants from all backgrounds – from the small start-up and established manufacturer to research teams from renowned universities – enter this competition to come up with innovative picking solutions. The aim is to bring autonomous storage and retrieval technology with the aid of artificial intelligence closer to reality. It is little wonder that the online giant from the USA is trying to advance the technology – after all, it is one of the world’s biggest retailers and has huge warehouses that would very much benefit from such a solution.
If the gripping problem is solved in the medium term, the robotic systems could be produced in higher numbers, which ought to bring the price down. For warehouse managers, the robots promise many good things: no time-consuming search for staff, no problem with absence due to illness, and 24/7 operation. The fact that a number of jobs will be put at risk, however, is a different matter, skills shortage or not.