MARLIN: A Cloud Integrated Robotic Solution to Support Intralogistics in Retail

TL;DR

This work presents MARLIN, a cloud-integrated service robot designed to assist intralogistics in retail by leveraging the semantic digital twin semantics of the K4R platform. It details MARLIN’s hardware stack, a 3D obstacle-detection pipeline, and a tractor-trailer navigation approach that surpasses the manufacturer’s baseline in confined spaces, as well as a planning workflow built on KnowRob and ROSPlan to accompany shelf replenishment. The semantic digital twin enables semantic queries and data-driven task planning for replenishment missions, while the K4R platform provides the cloud-based AI infrastructure and multi-agent orchestration. The experimental results across simulation, lab, and real-store settings demonstrate improved obstacle handling, navigation in narrow aisles, and end-to-end replenishment task execution, highlighting the practical potential of cloud-augmented robotics in retail with broader implications for scalable, fleet-based intralogistics.

Abstract

In this paper, we present the service robot MARLIN and its integration with the K4R platform, a cloud system for complex AI applications in retail. At its core, this platform contains so-called semantic digital twins, a semantically annotated representation of the retail store. MARLIN continuously exchanges data with the K4R platform, improving the robot's capabilities in perception, autonomous navigation, and task planning. We exploit these capabilities in a retail intralogistics scenario, specifically by assisting store employees in stocking shelves. We demonstrate that MARLIN is able to update the digital representation of the retail store by detecting and classifying obstacles, autonomously planning and executing replenishment missions, adapting to unforeseen changes in the environment, and interacting with store employees. Experiments are conducted in simulation, in a laboratory environment, and in a real store. We also describe and evaluate a novel algorithm for autonomous navigation of articulated tractor-trailer systems. The algorithm outperforms the manufacturer's proprietary navigation approach and improves MARLIN's navigation capabilities in confined spaces.

Figures (19)

• Figure 1: MARLIN: A mobile service robot for the support of store employees
• Figure 2: Sensor Processing Pipeline for obstacle detection on MARLIN. PC - Point Cloud.
• Figure 3: Visualization of the whitelist rules on a map of a retail Store
• Figure 4: Results of the Pointcloud Filter.
• Figure 5: Clustering, tracking and classification of three different objects in a laboratory environment.