With the proliferation of intelligent networks in industrial environments, manufacturing SME’s have been in a continuous search for integrating and retrofitting existing assets with modern technologies that could provide low-cost solutions for optimizations in their production processes. Their willingness to support a technological evolution is firmly based on the perception that, in the future, better tools will guarantee process control, surveillance and maintenance. For this to happen, the digitalization of valuable and extractable information must be held in a cost-effective manner, through contemporary approaches such as IoT, creating the required fluidity between hardware and software, for implementing Cyber-Physical modules in the manufacturing process. The goal of this work is to develop an architecture that will support companies to digitize their machines and processes through an MDA approach, by modeling their production processes and physical resources, and transforming into an implementation model, using contemporary CPS and IoT concepts, to be continuously improved using forecasting/predictive algorithms and analytics.