In the quest for renewable energy, Avangrid, a provider of clean energy solutions, implemented a novel AI application designed to help maintain wind turbines throughout the U.S.
The technology is called First Time Right Autopilot, which is a digital assistant to field engineers responsible for resolving difficult issues related to technical malfunctions.
Through the analysis of equipment documentation and error codes, the system helps resolve the problem, repair the equipment, and bring wind turbines back into service.
Digitizing technical knowledge for field technicians
Contemporary wind turbines have thousands of intricate electrical and mechanical parts working in harsh environmental conditions.
In case of any technical problem, engineers tend to go through hundreds of pages in operation and maintenance manuals.
The First Time Right Autopilot was designed by Avangrid and makes use of Amazon Bedrock services to create a compilation of all technical documentation.
The software constantly ingests complex machine manuals, maintenance records, technical drawings, and vendor fault code databases.
With the help of generative artificial intelligence, large volumes of technical information stored in legacy documents are consolidated into a central digital library.
Field engineers will get instant access to decades of technical information from one window only.
A centralized digital library will make sure that engineers always have access to technical literature.
Accelerating fault diagnosis inside turbine nacelles
If any kind of electrical and/or mechanical problems are detected by the internal sensors of the turbine monitoring systems, the system will automatically generate relevant diagnostic fault codes.
The specialists that work from significant heights in the turbine nacelles enter the fault codes in the mobile application.
Considering complicated fault codes in the generative models becomes possible through analysis of fault codes along with the troubleshooting history.
There is no need for the techs to search through the hard-copy manual since relevant suggestions will be offered immediately.
Thanks to this feature, there will be less guessing with complicated troubleshooting of electrical, hydraulic, and mechanical issues.
Thus, diagnostics become more efficient and save time.
Standardizing repair protocols across wind facilities
There are multiple wind energy farms in the US that have different types of turbines that are produced by different OEMs.
First Time Right Autopilot standardizes repair instructions for different hardware configurations and regional wind farms.
It provides very detailed instructions in steps that can be followed when repairing one particular turbine.
With such standardized instructions, technicians are able to perform their repair work just like senior engineers.
Standardized maintenance instructions ensure fewer human errors, keep turbine parts safe from any damage, and make sure that employees are safe.
Standardized instructions to repair turbines ensure the same quality of service for all operational renewable energy generation facilities.
Empowering field crews with mobile intelligence
The access to the artificial intelligence-driven generation tool is possible through mobile phones carried by field technicians who use them during their time in the field.
Questions about certain actions taken by components are posed using natural language, and the answers are delivered instantaneously in comprehensible technical language.
The mobility feature eliminates the need for the field technician to get down from the tower to refer to the documentation provided onshore.
Enhancement of the interaction between field staff and the onshore engineering department takes place through this procedure.
Increased efficiency of problem solving results in increased efficiency of the wind farm and in more renewable energy available for the power grid.
The mobile digital assistants have changed how field employees deal with renewable energy facilities, according to T&D World.
Outlining limits of generative maintenance tools
The implementation of the First Time Right Autopilot helps in making technical decisions, yet digital assistance cannot compensate for the actual physical labor done by humans.
The generative models entirely depend on the accuracy and completeness of the data supplied by the technical manuals and operation maintenance logs.
Humans are still required to perform physical inspections, do mechanical repairs, and verify the safety of electricity.
Unpredictable damage tovthe structure and wearing of the components will require engineering expertise.
Continuous data validation is necessary to make sure that the artificial intelligence model provides the correct diagnosis and troubleshooting advice.
Digital engineering assistants increase diagnostic effectiveness, but human expertise is required to assure wind farm efficiency.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
