Enel advances artificial intelligence powered drones

Enel Green Power North America and Raptor Maps are partnering on real-time identification of solar facility faults onboard drones.
Published: Thu 06 Sep 2018

The use of drones for line and other asset monitoring and management by utilities is growing, despite evolving regulation on issues such as out of line of sight use, proximity to people and restricted airspaces.

Coupled with their use, the implementation of artificial intelligence also is growing in enabling much of the monitoring process to be automated by comparing new data with old to detect changes.

The technology now takes another step forward with an agreement between Enel Green Power in North America (EGPNA) and the Massachusetts-based software provider Raptor Maps to implement and scale the use of drones and artificial intelligence for the management of solar assets.

“The partnership with Raptor Maps on this innovative technology goes beyond demonstrations and implements a ground-breaking solution in a responsible and scalable way,” boasts Rafael Gonzalez, Head of Enel Green Power North America.

Solar asset monitoring

The project is aimed to configure Raptor Maps’ existing machine learning/artificial intelligence software solution, Raptor Solar, which was developed for post-inspection analysis, and embed it directly into EGPNA’s drone hardware in order to allow real-time identification and classification of solar facility faults onboard the drone and streamline the detection-to-repair process from days to hours.

“By combining the new software with the technologies already implemented in our plants, we have the potential to increase efficiency of our inspections, yield more accurate results and work towards developing a more automated inspection process across all of our solar sites,” continues Gonzalez.

The project is being supported by the Massachusetts Clean Energy Centre (MassCEC) with an InnovateMass award, which was made to accelerate commercialisation of this technology.

Initially the project is implementing Raptor Solar across all of EGPNA’s solar assets, which comprise 22 plants with a combined capacity of almost 175MW. The technology will be able to simultaneously capture both infrared and high resolution colour imagery of these assets, perform post-processing at the source of the data, and deliver real-time analytics to assess the condition of the plant. This information will then be transmitted in real time to EGPNA’s Maintenance Management System, which will create and deliver a work order with actionable items to be evaluated by the site technician, all before the drone even lands.

“EGPNA is leading the industry in the scalable implementation of drones to streamline operations, maintenance and asset management. They are the perfect partner to push this technology to its full potential, and drive efficiency across the entire solar value chain,” comments Nikhil Vadhavkar, Head of Raptor Maps.

With this solution the companies are aiming to solve the data post-processing bottleneck that is common in drone inspections across solar plants to date by reducing the time and labour costs by creating a faster, more efficient process without the need to transmit large amounts of data over long distances. Once developed, the pilot programme is expected to be replicated across other renewable technologies.

Drones in energy

Artificial intelligence is just one of the new technologies being integrated into drones, and it is finding use not only in energy but for applications as diverse as building construction and crowd surveillance.

Indeed, according to its press note, EGPNA is scaling drone training and infrastructure to support the broad use of the technology across its development, engineering and operations groups for project location suitability assessment and construction monitoring, in addition to the use case above of streamlining operational maintenance activities.

Another is the concept being pioneered by Canadian public utility company Hydro-Québec of incorporating robotics into drones and such applications can be expected to grow so that the drone can not only reach a location and potentially capture imagery but also undertake in situ monitoring and analysis.

In its 2017 study on drones and robotics for transmission and distribution operations, Navigant Research estimates a more than six times growth in drones and robotics globally over the coming decade with the largest market in Asia Pacific. According to Navigant the three core drivers for the utility robotics market are speed, safety, and savings, and initiatives that demonstrates improvements in one or more of these is likely to be considered positively by regulators.

Nevertheless, the utility sector is still only a small player in the drone and robotics market. Taking Navigant’s estimate of the 2018 market at a little over $2bn, this is just 2% of the approximate $100bn market globally estimated by IDC. Of this over 90% is for robotics and less than 10% on drones.

“Groundbreaking improvements in [drone] technology has piqued interest by industries that operate in open or outdoor space, such as utilities, where inspection-related applications are key drivers of the industry.” Stacey Soohoo, Research Manager, Customer Insights and Analysis at IDC points out. “As policies change and governments work with vendors and end users to formulate regulation allowances, new opportunities and expanded use cases will come to light as their benefits are realized across all industries."