On-Edge Aggregation Strategies over Industrial Data Produced by Autonomous Guided Vehicles

Industrial IoT systems, such as those based on Autonomous Guided Vehicles (AGV), often generate a massive volume of data that needs to be processed and sent over to the cloud or private data centers. The presented research proposes and evaluates the approaches to data aggregation that help reduce the volume of readings from AGVs, by taking advantage of the edge computing paradigm. For the purposes of this article, we developed the processing workflow that retrieves data from AGVs, persists it in the local edge database, aggregates it in predefined time windows, and sends it to the cloud for further processing. We proposed two aggregation methods used in the considered workflow. We evaluated the developed workflow with different data sets and ran the experiments that allowed us to highlight the data volume reduction for each tested scenario. The results of the experiments show that solutions based on edge devices such as Jetson Xavier NX and technologies such as TimescaleDB can be successfully used to reduce the volume of data in pipelines that process data from Autonomous Guided Vehicles. Additionally, the use of edge computing paradigms improves the resilience to data loss in cases of network failures in such industrial systems.