Improving electrical power grid resiliency and optimizing post-storm recovery using LiDAR and machine learning

While many external factors influence resiliency, weather remains the single greatest threat to the electric power grid, and the impacts caused by significant storms can be long-lasting and widespread. When damage occurs, it is very costly to identify due to the vast size of electrical transmission and distribution circuits, which can span hundreds of miles. Pinpointing a failure in a circuit requires the expensive process of dispatching human teams to “walk the line” and physically inspect the circuit to identify damage. It is proposed that this problem can be optimized through automation, by leveraging flight vehicles, light detection and ranging (LiDAR) technology, and machine learning. The goals for this project are: 1) Investigate the feasibility of, and problems associated with, developing a system to remotely inspect electrical power transmission and distribution infrastructure with lidar. 2) Investigate the feasibility of developing an automated system to classify and detect damage to terrestrial transmission and distribution assets with lidar and artificial intelligence. 3) Develop a proof of concept of such a system, including a simulation of real-time lidar data collection and damage assessment