Generelles Ziel des Vorhabens SmartRecycling-UP ist es, über eine Automatisierung bisher manuell durchgeführter Prozesse eine signifikante Verbesserung der Recyclingquote und -Qualität für großstückige Abfälle (Sperrmüll, Bauschutt etc.) zu erreichen. Insbesondere sollen die technischen Voraussetzungen für eine Automatisierung der Vorsortierung großstückiger Abfälle geschaffen und demonstriert werden. Mit der neuen Technik soll die Effizienz des Sortiervorgangs, die Qualität der sortenrein dargestellten und für ein Recycling bereitgestellten Abfallarten und die Wirtschaftlichkeit der Sortierung bzw. des Recyclings signifikant gesteigert werden.

Duration: 01.01.2022 till 31.12.2024
Donee: German Research Center for Artificial Intelligence GmbH
Sponsor: Federal Ministry for the Environment, Nature Conservation, Nuclear Safety and Consumer Protection

Institut für Energie und Kreislaufwirtschaft an der Hochschule Bremen GmbH
Hochschule für Angewandte Wissenschaften Hamburg
Baljer & Zembrod GmbH & Co. KG
KreisAbfallVerwertungsGesellschaft mbH Minden-Lübbecke
Karl Siedenburg GmbH & Co. KG
Abfall-Service Osterholz GmbH

Project Management:

ZUG - Zukunft, Umwelt, Gesellschaft

Application Field: Logistics, Production and Consumer
Related Projects: SmartRecycling
KI und Robotik für eine nachhaltige Kreislaufwirtschaft (07.2020- 02.2021)
Robot systems for decontamination in hostile environments (06.2018- 06.2022)
Related Robots: ARTER
Autonomous Rough Terrain Excavator Robot

Project details

Figure 1: Rail-mounted crane from BZ (Photo: ASO, Osterholz)
Based on the knowledge gained in the conception phase, the final project goal of the implementation phase (SmartRecycling-UP) is to improve recycling processes by automating an electro-hydraulic excavator/crane for pre-sorting large-sized waste. Such machines are used today in various configurations. Figure 1 shows a rail-mounted crane from the manufacturer Baljer & Zembrod GmbH & Co. KG (partner in the implementation phase), which is used in a disposal centre of ASO Abfall-Service Osterholz GmbH (partner in the implementation phase) for the management of two deep bunkers of a handling facility. The machine's gripper arm and gripper are hydraulically driven. The machine itself travels electrically on a rail along the long side of the deep bunker. In this way, every corner of the deep bunker can be reached and both the shredder set up on the left edge of the bunker and various containers positioned near the bunker for contaminants can be loaded. The machine is operated manually by an experienced excavator operator.

In SmartRecycling-UP, the automation of both a mobile excavator and a rail-mounted crane is considered (Figure 2). In principle, the developed AI control and sensor technology should be largely independent of the end device (see also

The aim is for the automated crane to work with similar accuracy and speed as a trained excavator operator. This results primarily in two technical project goals:

Project goal I - Sensor technology: Sensor technology is to be developed with the help of which disturbing and valuable materials can be quickly and reliably detected in a mixture of materials typical for the recycling industry and an exact position determination can be carried out. Material detection can begin during the preceding sorting process and must be completed by the time the material is picked up (i.e. in <3 s in total). The position determination must be completed after the rough classification and before the start of the following grasping process (<0.5 s). Object classification must be performed as part of the coarse classification in the first 2 s. Due to the size of the gripper, the determined position may deviate by half of the gripper width across the gripping movement and by approx. 10 cm along the gripping movement. From a distance, objects from approx. 30 cm x 30 cm should be recognised. The aim is to recognise objects from 5 cm x 5 cm when the sensors on the gripper approach.

Project objective 2 - Autonomous manipulation: A commercially available hydraulic manipulator or crane is to be modified in such a way that it is capable of independently grabbing the disruptive and recyclable materials detected by the sensors sufficiently quickly and transferring them to separate containers/feed hoppers. In addition, the manipulator should be capable of efficiently and independently transferring the remaining material mix in continuous operation or transferring it to a shredder for further processing.


Smart Recycling - AI and Robotics for a sustainable circular economy

In the Smart Recycling project, in which the Robotics Innovation Center of the Research Center for Artificial Intelligence (DFKI) collaborates with the Institute for Energy and Recycling Management (IEKrW) and the University of Applied Sciences (HAW), as well as the Bremen waste disposal company Nehlsen GmbH & Co. KG was involved, developed a concept for the automated sorting of large-sized recyclable materials in the recycling industry. The SmartRecycling concept utilizes the most modern technologies from robotics, sensors and artificialintelligence in order to enable full automation of the pre-sorting of large pieces of waste and thereby increase the recycling rate.

Back to the list of projects
last updated 04.01.2024
to top