Leveraging Industrial Wireless Sensor Networks for Energy-Efficient HVAC System Operations

In industry, many processes require specific environmental conditions to guarantee product quality. Most of the time, these conditions, as well as the well-being of human workers, are ensured by Heating, Ventilation and Air Conditioning (HVAC) systems. In Germany, such systems account for at least 6.4% of all energy used by the industry. In order to reduce this amount while maintaining or improving the conditions for the human workers, we propose a Cyber Physical System (CPS) to control the systems based on the actual demand. To realize the CPS, we focus on three main points: (i) data acquisition through the sensor nodes of an Industrial Wireless Sensor Network (IWSN), (ii) reliable data transmission in the IWSN and (iii) closing the control loop through a decision framework. For the data acquisition we develop and evaluate a versatile wireless sensor node platform. The platform captures data with a high spatial resolution in factories through its low-cost approach. Together with a custom 3D airflow sensor, which has been specifically developed for this use case, the node provides valuable insight into the actual conditions inside a factory. To enable reliable data transmissions through the IWSN, we evaluate the usage of Multi-Connectivity (MC), which enables simultaneous data transmission through multiple independent network links. The benefit of using MC depends on several factors, such as the OSI layer on which it is implemented. Especially in the multi-hop IWSNs considered here, MC can either be implemented below or above the network layer. This decides if the MC approach is applied on a per-hop basis or end-to-end basis. To evaluate both approaches, we consider them analytically, through simulations and in a real-world deployment. Through this, general influencing factors on reliability, such as fragmentation, are extracted and discussed. The collected data is used as input for a simple control algorithm, which closes the control loop to the HVAC system. The complete CPS is exemplarily implemented in real industrial environments, showing the demand driven control approach. Finally, the monetary and environmental benefit of the CPS is calculated for projected energy savings of 5% and 10%.