Scientific curiosity and entrepreneurial spirit are a powerful combination to help innovations achieve a breakthrough: „MotionMiners“ represent both – four scientists who develop a solution to an automatic activity detection in intralogistics and make it ready for the market. The new initiative of Fraunhofer IML is promoted by Fraunhofer-Venture within the F-Days accelerator programme. With the concept to focus on the human being the project pays tribute to the activities of the innovationlab “hybrid services in logistics”. Dipl.-Inform. Sascha Feldhorst, employee at Fraunhofer IML and member of the initiative MotionMiners, explains how processes and ergonomics in logistics can be improved at the same time.
According to the current BKK (umbrella association of the German company health funds insurance) health report 2016 jobs in logistics show a particularly high sickness absence rate. Often, employees suffer from musculoskeletal disorders. Those developments show how important it is to improve the working conditions of humans in logistics. Our magic word is „MotionMining“. Behind that is an automatic activity detection as already known from sports or leisure. In logistics, however, the topic is new and largely unexplored. Expert interviews for a doctoral thesis revealed how much potential activity detection offers for intralogistics – be it that changes are evaluated in a pre-post comparison or technologies shall be selected that make changes possible in the first place. As a result process times in companies will shorten, workload of employees will diminish – a benefit for both companies and employees.
The innovative package we developed for this, kills two birds with one stone: collecting motion data in the warehouse and their following analysis. Mobile sensors, wearables and beacons help us to determine genuine process data in a first step, e.g. travel or picking times of employees during order-picking. Also, movements can be recorded which can cause debilitating disorders. Up to now, such data are determined manually – by monitoring with timer and clipboard – or simply estimated. The crux of such a process is: The order-picker is well aware of the monitoring and that can cause falsified results. To determine real – or more precise: realistic – data we will provide employees with wearables, e.g. an intelligent bracelet. By means of the wearables motions will be recorded on the one hand and information on the surroundings like temperature or light will be determined on the other hand. The additional use of mini transmitters, so-called beacons, helps to gain further knowledge about manual processes. With our solution all data will be collected anonymously and it is not possible to trace an individual person by means of the data. This processing has many advantages over existing camera-based solutions and is convincing for the representatives of both companies and workforce. For assessing the genuine data of the operational daily routine we developed a specific machine-learning solution. It allows to analyse the dataautomatically, i.e. by means of algorithms of machine learning. First, relevant motions from the data will be identified, afterwards they will be processed by means of a deep neuronal network (Deep Learning) and condensed for a process analysis. It is the objective of our technology to support employees and make their daily work easier.
Once more, our project is an example for the fact that logistics and IT go hand in hand today. In our team are experts for motion detection, database, sensor technology, deep learning and process analyses working together. Whereas our IT specialists are responsible for the registration and the computer-assisted analysis, our logistics professionals develop improvement proposals for the companies. At the moment, we are working at full stretch on a first use case. At LogiMAT in Stuttgart, the biggest yearly exhibition for intralogistics in Europe, this use case will demonstrate that the future belongs to MotionMining.
ABOUT THE AUTHOR
Dipl.-Inform. Sascha Feldhorst is scientific employee of Fraunhofer IML and expert for MotionMining and sensor technology. You can reach him on + 49 (0) 231 – 9743 590 or via email email@example.com
MORE ABOUT THE TOPIC
The webpage to the topic www.motionminers.com is currently being set up. For the time being, the information exists only in German.