In the improvement of robotic manipulator systems, open software platforms play an important role in control design of robots, especially where new techniques, as force control, want to be included. A good platform can provide both simulation functions that speed up the development of different algorithms and experimentation functions which include complex programming and huge data collection. This paper presents a new open software architecture for an industrial manipulator that permits easy implementation of model-based and sensor-based control concepts. In addition, it proves to be interesting for robotics research since it allows the integration of standard industrial components. The entire system has been successfully developed, implemented, and demonstrated for a robotic task which required the interaction between the manipulator and an unknown environment.