Optimising Intralogistics with AI
Interview on the AutoQML project
In its production facilities in Barntrup, KEB operates the in-house transport system AGILOX, which is designed specifically for intralogistics tasks. The AGILOX system is comprised of a swarm (union) of smart automated guided vehicles (AGVs), working collaboratively to transport items throughout KEB's warehouses.
Despite its sophistication, the system isn't immune to occasional operational or technical failures. When a vehicle stalls, it results in an immediate stop, necessitates human intervention, and ultimately leads to delays in operations. AGILOX system constantly generates data regarding vehicle status and orders, providing a valuable opportunity to utilize this data for further analysis.
In AutoQML – a project that develops solution approaches for linking quantum computing and machine learning – KEBs primary objective is to devise a machine learning solution capable of monitoring vehicle status and predicting potential failures. This aligns with KEBs larger objective of facilitating the broader transition to quantum computing in the future, by supporting research institutes with practical, real-world applications.
The machine learning solution to be developed in this project will be seamlessly integrated into the KEB Ecosystem to provide constant surveillance of the AGILOX system's health status, with the ultimate aim of enhancing intralogistics operations and minimizing downtime costs.
Khaled Al-Gumaei is Senior Data Architect IIoT at KEB. He is the project manager from the KEB side and one of the minds behind the project AutoQML. In the interview, he talks about his role and provides insights into the project.
What experience do you have in the field of quantum computing and artificial intelligence?
In my role as a data architect, my focus primarily lies in designing and developing data policies and infrastructure related to Industrial IoT. However, working closely with data analysts has augmented my understanding of AI and machine learning, particularly as they apply to Industrial IoT data analysis.
Regarding quantum computing, my exposure is currently more theoretical. Nevertheless, I am eager for the opportunity this project presents to apply this theoretical knowledge in a practical context, enhancing the utility in the Industrial sector.
What is your specific role in the project?
In this project, my primary responsibility is to create a data platform that encompasses the collection, transfer, storage, and management of AGILOX data, ensuring it is appropriately prepared and accessible for the AutoQML data scientists. Additionally, I bridge the gap between business and technology by translating the business requirements outlined by KEB’s domain experts into technical prerequisites for data analysis. Once the solution is developed, I will oversee its integration into the KEB Ecosystem, coordinating the implementation process to ensure seamless functionality.
What added value do you expect from the project content?
From this project, we expect several benefits. On the technical side, we aim to leverage the advantages of automatic machine learning (AutoML) techniques. These techniques will allow us to create AI models without extensive data science experience. Our goal is to discern complex patterns within the AGILOX system, thus optimizing our in-house logistics, simplifying maintenance, and importantly, reducing costs.
Beyond the immediate technical benefits, we believe our work also contributes significantly to the larger community. Our mission is to support research efforts focused on translating current industrial application challenges into problems that can be solved using quantum computing. In doing so, we hope to help drive the wider adoption of advanced technology and contribute to the evolution of digital innovation.
