In the Area Management Security Use Case of IOWN GF, the energy efficiency of computing infrastructure needs to be drastically improved for processing data in real-time image analysis of video streams from a massive number of surveillance cameras. As a solution to this problem, we conducted this PoC employing IOWN GF architecture such as Open all-photonics network (APN), data-centric infrastructure (DCI) with data-plane acceleration (DPA). That is, this PoC demonstrates the heterogeneous computing by Kubernetes (OpenShift container platform) enabling x86 Logical Service Node with x86 CPU, smartNIC, data processing unit (DPU), and GPU for the AI inference of video streams. In addition, performance evaluations have been conducted through Open APN to show that the use case can be achieved on geographically-distributed computing infrastructure without bottlenecks. As most IOWN GF members do not yet have a composable disaggregated infrastructure (CDI) system which is under discussion as a part of DCI Cluster in DCI-TF, this PoC report might help all IOWN GF members to see how the PoC can be started with a commercial off-the-shelf (COTS) server instead of a DCI system with CDI features. This also helps DCI-TF to learn about what type of composable Peripheral Component Interconnect Express (PCIe) device IOWN GF members are trying to use for the purpose of the use case reference implementation model (RIM) PoC.