The POWER Interview: Diving Into Data With Autonomous Inspection of Utility Assets
Companies offering the ability to collect data for use by electric utilities are finding their services in demand as the power generation sector embraces digitization.
It’s an industry that promotes innovation, as utilities need to track the real-time performance of thermal and renewable energy generation assets, and gain more insight about how their equipment is operating across their service territory.
North Dakota-based Thread is a software-as-a-service provider for electric utilities and the power generation sector. The company’s UNITI Workspace provides asset data management, workflow automation, and collaboration tools. Thread has developed a robotics platform to collect inspection data for utilities, and is helping automate data capture of utility assets across several sectors, including renewable energy, transmission and distribution, and oil and gas. Its customers include utilities such as Xcel Energy, and asset protection companies such as MISTRAS Group.
Josh Riedy, the company’s CEO and co-founder, has brought in investors such as Badlands Capital, Minnkota Power Cooperative—one of the company’s early utility partners—and Wonder Fund North Dakota, a group led by Kevin O’Leary of television’s “Shark Tank“ fame. The company is focusing its efforts on digitizing and modernizing energy assets, like wind turbines, and helping utilities in their transition to digital operations.
Thread’s product is a self-contained device with an on-board processor, paired with artificial intelligence (AI) algorithms and back-end management software. It autonomously collects data from energy assets, supporting customers in their deployment of robots and drones used to monitor assets. Customers include Constellation Energy, Otter Tail Power, Xcel Energy, and the U.S. Air Force, as the company works with that group to support security at Grand Forks Air Force Base in North Dakota, as part of a $1.2 million contract.
Riedy recently provided POWER with insight about his company’s work in the asset management space, and the importance of collecting and utilizing data to support operations of power generation equipment.
POWER: What are some of the ways power plant operators and generators of electricity can use data to make their operations more efficient?
Riedy: Electric utilities and operators depend on the reliability of their network. Dependable and available electricity is expected in most developed nations and interruptions can cause widespread disruption.
To maintain the resiliency of their networks, utility generators, transmitters, and distributors must maintain strict and effective inspection and maintenance programs. By repairing degrading assets and hardware, operators can sustain high uptimes for consumers. However, network reliability relies on rigorous inspection schedules, reporting, and data collection. Inspections of generation centers like wind and solar farms, and transmission and distribution lines, are usually done at regular intervals or when damage and operational interruptions occur. Often, only problematic infrastructure is recorded, leading to reactionary maintenance and repair. In turn, the evolution of health from correct operation to poor condition is largely unknown, leaving infrastructure vulnerable.
Want to learn more about how utilities and power generators are utilizing data to improve their operations? Read “Diversification of Power Generation Brings Greater Need for Data-Based Decisions,” part of POWER’s Special Report. And don’t forget to register to attend POWER’s Connected Plant Conference, happening May 20-22, 2024, in Houston, Texas.
Collecting and organizing data against each asset allows operators to keep infrastructure working by providing actionable insight into network health. Every image, video, scan, and annotation provides more detail for understanding risks. Including vegetation encroachment, weathering and corrosion of hardware, and environmental risks like storms and wildfires. Effectively collecting and aggregating data enables an analytical and objective approach to infrastructure maintenance including proactive and preemptive repair and replacement.
By taking a data-first approach to asset health and risk discovery, operators can reduce downtime, optimize repair schedules, and foster a safer, more efficient operational environment. Early hazard detection can lead to as much as 20% reduction in maintenance costs, translating to millions saved.
POWER: How can data be utilized to increase the performance of renewable energy plants?
Riedy: Wind-powered turbines are a growing source of renewable electricity generation. The cost of construction and maintenance of wind turbines is considerable, making regular inspection necessary.
Cloud-based technology, like UNITI, is designed to streamline the workflow so utility companies can manage energy generation infrastructure more efficiently and effectively through organizational and data management tools. Field inspections of equipment are the first line of defense against underperforming and deteriorating infrastructure. The data collected during an inspection provides the necessary insight required for maintenance and repair efforts. Proper data management grants engineers, operators, and managers the ability to generate insights into asset health and coordinate repair initiatives long before the point of failure. Cloud technology supports responsible data management with data organization and automation tools.
The collection of data and how it’s organized is the foundation for all systematic and analytical efforts to support high-functioning utility infrastructure. Without systems in place to support the data, inefficiencies are inevitable such as redundant truck rolls, reinspections, and laggard repair.
POWER: How can data be utilized to support decarbonization of the power generation sector?
Riedy: Today, energy companies are under pressure to implement technology that advances efficiency and productivity, risk mitigation, decarbonization, energy transition and changing consumer expectations. These considerations mean that operators are being pulled in a variety of directions, stretching resources thin.
Companies are implementing energy generation replacement projects that shift toward renewable energy technologies like solar, wind, geothermal, hydroelectric, nuclear, and more. Xcel Energy alone has promised to reduce carbon emissions 80% by 2030 from 2005 levels, and already operates more than 10,000 MW of wind power and more than 2,700 MW of solar.
To maintain this infrastructure, utilities rely on timely inspections to identify risks before they evolve. After data is collected, engineering and operations experts can review inspections and identify broken and depreciating conditions, such as delamination of wind turbine blades, lighting strikes, or impacts of weathering. Effective data use also supports the rehabilitation of equipment and the efficient use of resources when performing maintenance activities.
It is important to consider that inspection activities generate an insurmountable volume of data, from photos, videos, paper records, thermal imaging, drone monitoring, and more. Furthermore, inspection data is accompanied by weather records, vegetation encroachment, LiDAR and GIS, and even word-of-mouth knowledge sharing. This amount of data is difficult to review and make use of for rapid risk detection.
UNITI is based on an asset-first methodology using application specific taxonomy data structures developed in tandem with industry experts. This enables auto-associated contextual information for targeted and purposeful data collection and review. In turn, data management is reduced saving time for on-site staff, reviewers, and operation managers.
POWER: How can data support the electricity transmission and distribution network (can also compare and contrast Europe to the U.S., etc.)?
Riedy: Currently, utility inspections are reviewed asset-by-asset with the focus being individual components, poles, towers, or stations. Broad scale decisions are difficult to make when inspection data is so specific. As well, utilities regularly use contractors for both inspection and repair who have their own inspection protocols, leading to a further disconnect between data, decision, and action.
By aggregating data and incorporating legacy knowledge from experienced technicians, companies can begin to understand the evolution of grid health over time. In addition, companies enriching their inspection data with geographic metadata can group damages and repair activities, making response efforts more efficient. Geographically enabled data also has major implications during emergency events like wildfires and storms. Being able to quickly locate at-risk assets during critical times allows for faster response, mitigating the loss of power.
Digital ecosystems that help utilities integrate damage information like photos and inspection reports to decision-making architecture are equally as important as the data itself. Today, there is a serious lack of software to support turning discrete data collection into broad scale holistic grid health. In the UK, the Resilient Electricity Networks for Great Britain (RESNET) project aims to “develop models relating changes in the severity of weather-related events to failure rate of individual network components.” Yet, software solutions that support such initiatives are largely missing.
Understanding how and where grid infrastructure is in jeopardy is critical to building redundancy in both physical equipment and operational response mechanisms. While data is the foundation to systematic change, that information needs to be recorded effectively and intuitively so that it can be used to support network resiliency programs.
The UNITI Workspace is used by a growing number of U.S. utility companies to manage their assets, including generation, transmission, and distribution networks. UNITI has tools that support field inspection directly from mobile devices, automated robotic drone flight, and for integration with workorder management tools (EAM) and asset performance management (APM) software. Contractors and staff use UNITI to ingest data into the same system, eliminating confusion and building trust in asset data. Already, Thread clients are reporting increased asset longevity by using event-based inspections and efficient data collection. UNITI tools help clients realize a proactive monitoring structure that clears the way for preemptive maintenance action. In some cases, asset lifespan is projected to double from 15 to 30 years.
POWER: How is data being used to support decentralization of power generation, and are there better ways it can be utilized to make power production and the power grid more efficient?
Riedy: Decentralized energy generation is growing. In Brazil, by March 2023, more than 1.8 million renewable distributed power generation systems had been installed, equating to 19 GW (of generation capacity). Net-metering policies are largely seen as the catalyst.
The current grid network in many countries is built with similar architecture—generate the power in a few locations and transmit that power out to customers. Since critical infrastructure, such as hospitals, transportation, and food storage and production all depend on uninterrupted power, there is significant regulatory oversight for centralized generation systems. Supervision helps provide equitable access to power, minimize downtime through CAIDI/SAIDI metrics, and standardize reporting for utility companies. It also means that monitoring, inspection, and repair, is easy to manage and coordinate by the responsible company. However, the entire line from generation to consumption must remain intact to ensure the stability of electricity. Faults in connectivity anywhere along the path can result in service instabilities or outages.
Decentralized energy production has the potential to significantly reduce both the risk of outages, and the environmental impact by bringing production and storage to a local or micro level. In fact, under the right conditions, each individual home could install solar and battery storage to become self-sufficient or even generate excess power for neighbors. Decentralized design does have drawbacks though, such as initial setup costs, difficulty in regulation, monitoring and inspection, and the responsibility of repair.
Net metering, or the idea that excess decentralized power is fed back into the grid, adds additional complexities surrounding management and maintenance. Utility operators are faced with integrating modern power production into archaic systems not designed for bi-directional flow. These new systems may benefit from equally new technologies such as AI/ML, drone and robotics monitoring, and statistical modeling.
Still, no matter the source of the data monitoring software like UNITI Workspace allows enterprises to create reports detailing asset and network health. Reports can be used to help audit the decision trail, standardize operational intelligence, and share with governing bodies.
POWER: What can data do, if anything, to make power generation more equitable and sustainable?
Riedy: The inspection of assets is a manual, human-led activity that requires an abundance of person-hours to perform. If collected data isn’t properly recorded and organized in a way that enables retrieval and modeling, then identifying deteriorating conditions can’t be done. The result is that assets don’t receive the due care and attention necessary for proper function. Existing workflows also lack asset-to-asset damage comparison hampering the ability to prioritize restoration initiatives.
Because of these missing mechanisms, it’s not uncommon for utilities to run assets to failure. Typically, this means distribution poles and transmission towers are only replaced once they become non-functional. In some cases, the same is true for wind generation turbines.
The Utility Analytics Institute said, “Until now, utilities have had limited ways to anticipate and mitigate these equipment-related issues, leaving electric utilities with few alternatives other than to wait for equipment to degrade to the point of failure.”
To compound problems, inadequate grid infrastructure is usually geographically associated with underprivileged areas and lower socioeconomic classes. According to American Progress, “[Energy poverty] disproportionately affects people living in rural areas, as many energy-deficient communities are located in areas that are too remote or impoverished to attract investment for centralized grid access.”
Our ability to defend against failing hardware, build a sustainable future, and bring equitable access to electricity, depends on resilient infrastructure and adequate preservation of assets. Even for non-renewable sources of energy production, the grid must be always in healthy operational status to transmit and distribute electricity to consumers.
Generating useful and exploitable inspection data for proactive conservation of transmission and distribution is the first step in generating a dependable grid.
POWER: How can data be utilized for predictive maintenance and remote monitoring of power generation assets?
Riedy: For commercial wind power generation and solar farms, many inspections are performed using drones. Whether these inspections are performed by a Drone Service Provider (DSP) or in-house, most are manually flown. Consequently, repeatable flights and temporally comparable images are challenging to achieve. Inconsistencies like image location, distance, size, and resolution leads to cumbersome and subjective comparisons of disparate data and missed opportunities to capture problems before failure. Moreover, the quality of the mission depends on the experience of the pilot.
The right software tools present the opportunity to advance inspection through automating data collection and management. Together, these aspects enable an analytical and statistical approach to maintenance and repair routines. By modeling failure types, health depreciation rates, and downstream impact, utilities have an opportunity to forecast risk. Timing repair allows utilities to be efficient, not only with parts and labor, but also planned outages and operational interruptions.
When it comes to wind turbines, monitoring blades for damages in efficient ways can be one of the most challenging tasks with site managers, technicians, and fleet engineers. Hidden defects that often are not visible to the human eye are even harder to detect. One innovative solution is to install IoT devices like Sensoria, a 24/7/365 blade monitor developed by MISTRAS, to continuously detect and report blade integrity.
UNITI Workspace provides sharing and collaboration tools that link operation and engineering teams like never before. UNITI provides all the necessary components including automatic data-to-asset organization, cooperative discussion tools, and exporting and 3rd-party integration of data, imagery, and insights, resulting in a more holistic approach to monitoring.
Together, always-on IoT devices like Sensoria, paired with the UNITI Workspace offer a place for monitoring and data intelligence to live together. “With Thread’s support, MISTRAS can achieve its goals in part to validate inspection findings and alarms for the blade maintenance program, helping customers transition from reactive blade repairs to proactive blade maintenance. By corroborating Sensoria’s findings, we’re able to estimate the rate of degradation of our blades which makes budgetary planning easier” says the VP of Engineering and Product Development at MISTRAS.
With better data, and more people working together, utilities can start building a network-wide understanding of grid health and asset lifecycle to coordinate maintenance efforts well before the risk of failure.
—Darrell Proctor is a senior associate editor for POWER (@POWERmagazine).