Back to search

BIA-Brukerstyrt innovasjonsarena

SmartWing Infrastructure Failure Tracking and Emergency Response

Alternative title: SmartWing dronebasert beredskap og feildeteksjon

Awarded: NOK 13.7 mill.

The SWIFTER project shall develop a completely autonomous long-range drone-based emergency response system that enables grid operators to localise and diagnose critical faults more safely, quickly, and cost-effectively than current methods. Critical failures (i.e. those that interrupt grid operation), are an expensive problem for grid operators worldwide, both in terms of loss of reputation and income from consumers as well as fines for grid downtime. Distribution System Operators (DSOs) in Norway estimate that their annual costs related to such downtime can make up to 1-2% of revenue. These incidents also have flow-on effects for consumers such as increased energy costs and reduced provider satisfaction. When critical failures occur, it is therefore crucial for the grid operator to be able to localise and diagnose the cause of the failure ASAP, such that response operations can be coordinated quickly and cost-effectively. Today, grid operators most commonly deploy field personnel by foot, vehicle, or manned helicopters as the first response following an incident, and faults are identified manually. These methods are slow, time consuming, expensive, and expose personnel to unnecessary risks. Autonomous drone-based systems could directly address all these concerns, but today's commercial solutions are hamstrung by limited flight ranges and unreliable AI for automatic fault detection - preventing wide-scale use for critical operations. SWIFTER will enable KVS Technologies, in collaboration with SINTEF, Elvia, and Agder Energi Nett, to develop a completely autonomous system based on a long-range drone platform that will enable fast and cost-effective localisation of critical faults anywhere in the network at any time. By targeting only the most critical faults and by employing a multi-sensor approach, SWIFTER aims to improve on today's state of the art in terms of reliability and poor weather performance, thereby enabling wide-scale deployment of such systems by DSOs. In the project to date, KVS Technologies has designed and built a new drone platform specialised for long-range missions and BVLOS operation. This drone platform is now in production and flying commercial missions. KVS and SINTEF have designed a new multi-sensor suite tailored for detection of power lines, masts, and critical failures under poor visibility conditions, which will commence flight testing in Q4 2021. In collaboration with Agder Energi Nett, a remote segment of the grid network that was slated for replacement by underground lines has been used to re-create a variety of critical failures such as fallen trees and lines, broken lines and masts, missing components, and branches across the lines. Aerial sensor data capturing these failures is being utilised by KVS Technologies and SINTEF to develop an automated pipeline for failure detection and classification. Finally, this project so far has facilitated good dialogue between the project members on the usage and current limitations of various sensor data and automated tools employed by network operators.

The goal of this R&D project is to develop a robust and autonomous drone-based solution for the detection and real-time reporting of critical failures in the overhead line power grid. The project will develop and integrate a new sensor suite enabling failures to be located and diagnosed in low light and under a broader range of weather conditions than is possible with helicopters today. The sensor suite will provide the necessary sensor data for onboard AI to detect faults and report feedback to the grid operator. Failures often occur in remote areas with poor mobile network coverage, so it is not feasible to livestream high-definition video data and images to a remote human operator to detect failures. The drone must therefore be able to analyse sensor data automatically on board, so that failures can be reported in real-time to coordinate response operations. The sensor suite and onboard AI will be developed to be compatible with a long-range fixed-wing drone platform, enabling efficient inspection missions covering up to 100 km of overhead line during one mission.

Publications from Cristin

No publications found

No publications found

No publications found

No publications found

Funding scheme:

BIA-Brukerstyrt innovasjonsarena