GPUs and programmable data planes have gone through an enormous evolution in the past years. GPUs can be used for modeling the real-world environment accurately, while programmable data planes can monitor the network in real-time and implement novel packet processing and decision logic. In this paper, we investigate how these two distant technologies can be combined to implement efficient coordinated multi-point (CoMP) transmission in an indoor environment. The proposed system implements the concept of dynamic on-off switching (DOOS) among various radio transmitters, relying on two information sources: 1) radio propagation digital twin based on real-time simulations of radio channels in the accurate 3D digital representation of the real industrial environment and 2) traffic load collected by the data plane for each transmitter. The proposed DOOS method implemented as a data plane algorithm dynamically selects a set of radio transmitters for each receiver by mixing information provided by the digital twin and the in-network traffic load measurements. The proposed method has low computational complexity and reduces the number of actively used radio transmitters. The proof-of-concept implementation of the proposed system has been validated in simulations as well as with measurements. The method achieves 69% energy saving for the radio compared to the default CoMP transmission and reception.