Low-cost testbed setup

With the widespread adoption of IoT infrastructure, contemporary smart applications transcend the confines of a singular domain. Whether within a home environment’s boundaries or spanning a city-scale infrastructure, multiple smart applications, and their associated services collaborate on a shared sensing platform to cater to users’ needs. The primary motivation behind this shared infrastructure is to minimize installation and maintenance costs, but it also facilitates rapid cross-domain inference. This means that a common data set, such as room temperature and CO2 density, can simultaneously inform occupancy estimation services, feeding into HVAC systems, surveillance systems, and smart resource management systems concurrently.

To achieve faster decision-making and automation, it is crucial to preprocess sensory data at the network edge. However, current simulation platforms fail to accurately capture the real-world dynamics that impact the performance of edge-based sensing and data processing solutions within a shared IoT infrastructure. We therefore develop a testbed to assess the performance of such edge-based solutions, specifically in supporting ambient living. The main objective of this testbed is to support the adaptive sensing and edge-based processing of gathered data. The aim is to reduce computation, communication, and storage overhead in the continuous automated monitoring of infrastructure while maintaining the quality performance of stakeholder applications. The testbed includes diverse sensing nodes deployed within the Academic Building. Emphasis has been placed on accommodating multiple applications with varied Quality of Service (QoS) requirements, all utilizing a shared sensing infrastructure. This approach aligns with the futuristic design goals of next-generation IoT systems.

Further details can be found here: “Experience: Developing a Testbed for Ambient Sensing and in-Network Data Processing”