This paper presents the prototype implementation of an advanced automation architecture for electrical railway systems, designed to operate them as cyber-physical systems, such as smart grids. This REM-S architecture, from the literature, is a distributed system coordinating rolling stock, electrical substations, sideway resources within the constraints, and objectives (overall energy demand, power consumption, and cost optimization) dictated by the control center. The prototype software suites presented here are the REM-S Offline Suite and the REM-S Online Suite. Among others, they consist of an application for day-ahead optimization, which is performed by the control center for the next day, and another application for minute-ahead optimization, which is continuously performed during rail operation on so-called intelligent substations at time intervals of, e.g. 15 min. The REM-S Offline Suite can be used for dry runs of test scenarios before running the aforementioned applications in combination with others on the participating trains, substations, and so forth, as part of the REM-S Online Suite within a railway system in real time. The REM-S Online Suite, which is an implementation of a distributed optimization, was validated in a field test, performed in a suburban 3-kV dc railway line in Malaga, Spain during real rail operation where the objectives within the constraints given by the control center have been successfully reached. Besides, the software architecture of both software suites also selected algorithms that are presented for a better understanding of the overall communication and distributed real-time optimization processes.