Critical Infrastructure inspection is nowadays a costly process, as it requires operators reviewing infrastructures to check distances of thousands of kilometres which are often placed in difficult access areas, which makes it expensive in terms of time and cost process overall in Utilities, Construction and Services sectors.
UAVs are equipped with many types of sensors and gather information in real time. To manage this volume of data in an operative manner, with a fast response, and in a cost-efficient way is hard without the appropriate framework, so UAV applications need to optimize data management and analysis to decide what data to process and analyse locally and what data to transmit, assessing network constraints to get timely responses.
Today only 5% of data gathered by Unmanned Aerial Vehicles (UAV) is processed due to bandwidth and connectivity limitations, and still Utilities and Services companies spend millions of euros inspecting infrastructures manually, from power lines to pipelines, in hard-to-reach and dangerous locations, without having high-quality detailed images of infrastructure with some kind of damage, corrosion, among others.
Gartner predicts that Industrial inspections will represent 30% of the commercial drone market by 2020, and Edge computing is now recognised as the facto execution environment for these rich IoT scenarios, while including state-of-the-art technologies such as Computer Vision which has been recently largely enabled due to advances in Convolutional Neural Networks and availability of the adequate hardware for the computations at Edge/Cloud environments.
The convergence of this scenario, and current needs of Critical Infrastructure Owners/Operators, as well as Managed Surveillance and Inspection/Monitoring companies, potentialities the creation and development of a solution as E-FLY, to build a distributed UAV, Edge and Cloud AI image processing framework for the analysis of infrastructures, with Edge Computing acting as the ground-breaking technology that allows the ability to exploit data gathered from UAVs.
E-FLY will be proven on a pipeline infrastructure inspection pilot case, which will demonstrate the benefits gained from the application of E-FLY to UAV image processing to move from manual operations to fully automated ones through UAV, Edge and Cloud powered process, such as:
Cost and time savings
Less time on-site due to rapid data collection
Access to difficult to access areas
Multiple datasets in a single one flight (visual inspection, orthophotos, etc)
Real-time data processing
Automatic detection of incidents
Simplified maintenance task
Assessment of evolution of incidents
Anticipation and early warning of possible anomalies