ROTATOR

Empty and out-of-stocks are a major problem in intralogistics, especially in retail.
A much-cited study analysing out-of-stocks (OoS) in over 600 markets worldwide concluded that a retailer loses about 4% of turnover as a result (cf. [Gruen 2002]). The OoS rate is of course partly caused by errors in forecasting and ordering, but also to a quarter by so-called shelf OoS - cases where the goods are in the store but not on the allocated shelf space. If a large supermarket/hypermarket - of which there are about 1000 in Germany - generates a turnover of about 30-40 million € per year and one assumes a trade margin of 20% on average, the market directly loses about 80,000-100,000 € in contribution margin per year through Shelf-OoS alone. In this context, it is particularly important to recognise sell-outs during promotions/second placements (promotion tracking), because experience shows that the OoS rate is much higher there and retailers lose additional advertising cost subsidies if the goods are not placed.

The aim of the joint project is therefore to detect shelf OoS (Out of Stock) in a timely and autonomous manner using a mobile platform (e.g. cleaning robot) whose OoS unit records the environment in three dimensions. Human-machine interaction is of central importance here, because the success of the machine requires interaction and cooperation with humans, be it in cases where customers stand in the way of the robot or cover shelves, or in order to achieve that employees accept the robot as an assistance system.
ROTATOR is pursuing a transdisciplinary approach within the focus area of production/logistics; research results are being developed for 3D information acquisition (sensor technology with high spatial and temporal resolution for moving objects and scenes), 3D information processing (real-time processing, mapping, pattern recognition and classification of OoS and people). The human factor is explicitly addressed in acceptance and requirements analyses, both in terms of considering the needs and capabilities of stakeholders and defining the interaction functionality of the machines. The vision of the project partners is that with the OoS unit developed in the project with 3D mapping and the human-machine interaction procedures, mobile robots will be able to perceive their environment so smartly that they could then also clean safely in customer operations during opening hours, report OoS to the head office "on the side" from 2020 and perhaps also refill the shelves independently from 2025.