In distributed manufacturing environments, Auto-mated Guided Vehicles (AGVs) equied with visual camera play a crucial role in automating material handling and optimizing production efficiency. Detecting anomalies during AGV operation is crucial to prevent potential malfunctions that could disrupt industrial processes. However, anomaly detection is challenging due to privacy concerns and the heterogeneity of data collected by AGVs across different factories. While sharing data across factories can improve the generalization capabilities of models, this can lead to overgeneralization in reconstruction-based anomaly detection, where the model reconstructs both normal and anomalous data too well, reducing its ability to detect anomalies. To address this problem, we propose FedCali, a federated learning framework that balances generalization and specialization across AGVs monitoring different manufacturing processes. Our proposed Gradient Guiding Mechanism (GGM) selectively aligns local model gradients with global knowledge only when necessary. This allows local models to retain their unique characteristics while benefiting from shared insights. Experiments with the MVTec dataset show that FedCali improves both reconstruction quality and anomaly detection accuracy, achieving higher AUROC scores and lower losses compared to baseline methods. This shows that FedCali is able to effectively process various manufacturing data collected by AGVs while maintaining data privacy.

Software-Defined Network (SDN) is the new paradigm in the computer network architecture. The concept is based on the decoupling of the data plane from the control plane. Such decoupling creates the possibility for central traffic management, hence offering the potential to improve the network’s performance and monitoring capabilities. We use those capabilities to perform extensive testing of the communication system for the fleet of Automated Guided Vehicles (AGV). First, we present the configuration of our testbed, the methodology of performing such testing, and a framework for experiment design. Finally, we present our results based on the presented approach.

Automated Guided Vehicles (AGVs) are a key component of many modern industrial systems. AGVs are supposed to communicate with each other in real time using wireless networks. In this article, the advantages and disadvantages of the ZigBee wireless network related to the control of AGVs are considered. We analyze the performance of the ZigBee network programmed with both C# and Python libraries to control ZigBee devices. The throughput and signal strength are presented and discussed depending on the transmission speed of the serial port, the payload size, and the presence and distance from the obstacles. The results of the experiments show the effective values of these parameters, the methods of using C# and Python, and the reliability of the throughput up to a certain point in network devices.

The use of Artificial Intelligence (AI) to support the Automated Guided Vehicles (AGV) that are used by industry poses a number of challenges that are specific to smart internal logistics systems that are necessary for agile manufacturing. On the one hand, it might seem that experience with the autonomous navigation system that are used in autonomous vehicles can be easily transferred to AGV. However, in this paper, the authors highlight specific problems that are associated with the navigation system of AGV, which has to reflect its operation in an industrial environment with high level of interaction with other production systems and human staff. On the other hand, it may seem that the wealth of experience from using AI in smart manufacturing can be easily transferred to the use of AGV. However, the authors show that although AGV are production tools, the challenges that are associated with the use of AI can significantly differ from other smart manufacturing areas. The number of challenges that are specific to use of AI for AGV is also discussed. This paper systematizes these challenges and discusses the most promising AI methods that can be used for the internal logistics systems that are based on AGV.

Automated Ground Vehicles (AGVs) use deep-learning-based vision systems to perceive the surrounding environment and extract relevant information about it. Although deep learning models offer high capabilities, they require large amounts of data to be properly trained and tested. Testing is especially important when off-the-shelf models are used by the AGVs - to examine whether they can meet the demands of complex environments such as the production halls of automated factories. One area of such perception algorithms is object recognition. To test such systems, we propose a solution based on ArUco fiducial markers used for automatic labeling of objects. Our solution can be used to test deep learning systems in real time directly on a robot. Our solution requires minimal interference with the environment and additional infrastructure - the desired objects only need to be marked with a marker printed on a home printer. Therefore, the presented testing procedure can be used for testing of AGVs in real-life environments during a real ride from an actual robot perspective. Data gathered during the online testing can be used for the offline comparison of the accuracy of different deep learning models. Although we focus on the online and offline testing in our study, we also incorporated a marker masking procedure. Therefore, the resulting datasets may also be used for training.

The article presents the results of a study to verify the possibility of discovering the type of work performed by a monitored object. During the research, the monitored object was an AGV streaming data about its current state. Each value representing the state of the AGV was transmitted in a separate stream. The data transmitted could be at different frequencies for each stream. The goal was to verify the possibility of discovering the type of work performed by the AGV on the basis of data that was generated only by the monitored object (without data from external systems). In the course of the work, a mechanism was developed to identify the beginning and end of the work performed by the AGV, as well as a way to aggregate the values characterizing the work performed. The set of characteristics of the work was selected in a manner that allowed easy interpretation by AGV fleet managers. Discovery of the type of work performed was done using two clustering algorithms: KMeans++ and DBScan. The set of features analyzed by the algorithms was selected experimentally. The results obtained with the two algorithms were compared. The identified work types were used to create work profiles, characterized by feature sets and appropriate value ranges.