IIC Journal of Innovation 5th Edition | Page 29

Edge Intelligence: The Central Cloud is Dead – Long Live the Edge Cloud! (4) Interaction between ECNs, user information synchronization, control and orchestration pre-identify the faults which will reduce the maintenance expenditure. (2) Security in preventive maintenance In some scenarios, the ECNs need to work in a cooperative way. For instance, at night, lamps need to increase their brightness when vehicles and pedestrians approach to improve safety and decrease the brightness when they leave to save energy. This process is expected to be continuous, especially the handover between two ECNs. Based on the headings, velocities and positions, an ECN can predict the next ECN that the vehicle will pass and remind the latter to get ready for handover. Malfunction reporting and data uploading should be authenticated and encrypted. On- site engineers must also be authenticated, then authorized to access to maintain and repair. Indoor Location Tracking (1) Network bandwidth High local network bandwidth (with reasonably low latency <100 MS) and provisions for backhaul to a local processing node with sufficient compute to allow the calculations necessary will be needed. Smart Elevator In addition to cloud offloading and privacy, and interaction between ECNs functionality, as described in the smart city lighting use case, the smart elevator use case would require: (2) Provisioning The ability to add this capacity seamlessly within existing and future networks should be provided. (3) Edge compute power (1) Diagnosis and predictive maintenance The local processing node should have a level of processing capability equating to, at time of publication, a server class machine. This can be achieved either by dedicated hardware or better via containerization within spare capacity on existing general purpose servers, or if future developments allow on an embedded compute node. Elevator safety is paramount, and fault detection and maintenance is critically important. Currently, on-site maintenance requires highly trained personnel using advanced tool sets to diagnose and repair, which leads to extra expense in staff training and equipping. Moreover, it often takes a considerable amount of time to identify the fault and find the corresponding soluti on. Lone Worker Safety (1) Power efficiency Autonomous, accurate and timely diagnosis helps locate the malfunction once it happens; while predictive maintenance gives the alert before malfunctions occurs. Both of these mechanisms will help to improve the safety of the elevator and help IIC Journal of Innovation The footprint required for the location equipment is bulky due to the power/antenna requirements of current systems as well as the power requirements for maintaining GPS operation. - 27 -