| |

International Research Project Federated Learning for Robot Picking (FLAIROP) Successfully Completed

New AI methods improve picking by robots

August 3, 2023

International Research Project Federated Learning FLAIROP Successfully Completed

The FLAIROP (Federated Learning for Robot Picking) research project, which was sponsored by the German Federal Ministry for Economic Affairs and Climate Action, was concluded at Festo in Esslingen-Berkheim. At the event, all project participants – as well as interested public – were on site or presented their results live from Canada.

Over the past two years, Festo has been conducting joint research with the “Karlsruher Institut für Technologie” (KIT) and partners from Canada (University of Waterloo, Darwin AI) to make picking robots more intelligent using distributed AI methods. To do this, the partners investigated how robots can learn from each other without sharing their training data. This approach – called Federated Learning – allows to develop more robust and efficient AI than it would be possible with data from just one robot – without handing out sensitive company data.

“We are proud that we have succeeded in showing that robots can learn from each other without sharing sensitive data and company secrets. This protects our customers’ data and we also gain speed because the robots can take over many tasks faster this way. For example, the collaborative robots can help production workers with repetitive, heavy and tiring tasks,” says Jan Seyler, Head of Advanced Development Analytics and Control.

Jan Seyler, Head of Advanced Development Analytics and Control.

“We have developed a universal, simulation-based data set that we can use to train autonomous gripping robots in such a way that they are able to reliably grasp items that they have not seen before,” explains Maximilian Gilles from KIT. In the future, the Federated-Learning System will be further developed so that the platform enables different companies to train robot systems together without having to share data among themselves. This can increase the acceptance of such systems in practice.

Maximilian Gilles from KIT

Federated Learning revolutionizes picking through robotics

Federated Learning is a machine learning technique to create privacy-preserving AI applications. Instead of sending the training data of the robotic arms in the picking cells to a central server to train the model there, with Federated Learning, the training takes place at many different locations. The locally trained models are then sent to the central machine learning server so that the sensitive training data does not leave the data provider. Nevertheless, Federated Learning enables learning across data silos by aggregating the distributed models and ultimately enabling highly accurate and data-driven prediction of object recognition and grasping point detection.

The robot arms in the picking cells are equipped with cameras to visually detect the items in front of them. Based on the camera image, the robot arms automatically recognize the different items and select a suitable gripping method. Due to the variety of items in a warehouse, this is a complicated task and large amounts of data are needed to achieve reasonable results. Creating such large amounts of data is time consuming. With data collected from picking cells in different organizations, it was possible to improve the grasping point detection of the cells.  

During the project, a total of five autonomous picking stations were set up for training the robots: Two at the KIT Institute for Materials Handling and Logistics Systems (IFL) and three at Festo SE & Co. KG based in Esslingen am Neckar.

Project completion: Now it’s time for practice

At the final event, Festo focused on the usability of the results. Jan Seyler: “We are showing which Festo products it can be incorporated into.” The research results will now be published and can be freely used by all interested parties in initial pilot projects.”

More Information

FLAIROP

Related Product

Blind Bin Picking with GripperAI

The software GripperAI from Festo enables “blind bin picking” and thus autonomous handling solutions with almost any robot, cobot or Cartesian gantry. The company ibk Ingenieur Consult GmbH is offering an online configurable robot cell as an entry-level model. This could become the patent solution for many gripping tasks, including customised designs and integration services.

Which user of robot, cobot or Cartesian gantries doesn’t dream of being able to recognise and grip unfamiliar objects or objects that have not been taught in, including automatic gripper changes. The GripperAI from Festo might just make this dream come true in the not too distant future. With conventional solutions, you either have to teach in using CAD or manually teach in the robot’s movements.uding customised designs and integration services.

Related Articles


Latest Articles

  • Transforming Manufacturing Processes with AMRs: Boosting Efficiency and Adaptability

    October 2, 2023 In today’s rapidly evolving business landscape, companies are constantly seeking innovative active solutions to optimize their manufacturing processes. One technology that has gained significant traction is the implementation of Autonomous Mobile Robots (AMRs). These versatile machines offer a wide range of benefits, empowering companies to transform their manufacturing operations and stay ahead… Read More…

  • How RFID Technology Is Expanding Manufacturers Traceability Capabilities

    October 2, 2023 Radio Frequency Identification (RFID)Technology has bee making great strides in expanding the capabilities that manufacturers have in Traceability. Due to its industry-proven, robust, short, and long-range design to identify, track, and trace products and carriers, RFID offers solutions beyond optical barcode reading. It can also be easily integrated into automation platforms using… Read More…


Featured Article

Guillevin Strategies and Solutions: Enhancing the Customer Experience With a Comprehensive Website

Guillevin Strategies and Solutions: Enhancing the Customer Experience

In today’s customer-centric business climate, creative solutions are needed to provide customers with the seamless experience they’re looking for. By offering a comprehensive, easy-to-use website that meets the diverse needs of its customers, Guillevin meets this challenge. 

With the aim of building lasting relationships with its customers, the website serves as a platform that not only simplifies the purchasing process, but also contributes to the success of both short- and long-term partnerships.

Read More


Products

  • New Endress+Hauser Pulsed Radar Level Sensor from AutomationDirect

    October 2, 2023 AutomationDirect has added the Endress+Hauser Micropilot FMR10 series pulsed radar liquid level sensor to their growing lineup of liquid level sensors. This new sensor is a downward facing, non-contact sensor that emits electromagnetic pulses and determines the level based on the time it takes for the pulses to be reflected back to… Read More…

  • PBC Linear Shafts, Bearings, and Shaft Supports from AutomationDirect

    October 2, 2023 AutomationDirect has added PBC linear shafts, bearings, and shaft supports to their power transmission product category. PBC linear shafts are precision ground to provide an optimized surface for plain and ball bearings and are ideal for high-precision linear motion applications where tight tolerances are needed. The shafts are available in carbon or… Read More…