This work presents a platform based on edge computing for the implementation of monitoring algorithms in complex industrial processes, combining safety, performance and scalability. Only free software was used and the connection of the industrial computer with processes was made via OPC UA communication standard, very common in industry 4.0. Two case studies were worked out to explore the features of the platform. First, multivariate process monitoring was applied to a pilot plant, showing steps towards training and remote monitoring. Secondly, increasing volume of synthetic data was provided by an OPC server to evaluate three performance metrics of a single node and to decide about inclusion of new nodes to share computational burden. The results achieved show the proposed environment can handle computational tasks of increasing complexity for general applications in the industry.
