In the wake of a boom in renewable energy, energy companies must scale up in order to meet increasing demands. This has made construction timelines and the reliability of energy production more imperative than ever before.

Here’s the thing. The average solar construction project is delayed by about 20% with the consequences hovering around $2M in costs. For example,  according to the EIA (US Energy Information Administration), about 20% of the solar photovoltaic capacity that was planned from January through June 2022 didn’t go online as scheduled.

More often than not, current methods to meet construction deadlines fall short because they are usually done manually, inconsistently, infrequently, and mainly rely on third-party vendors. 

Manual, ad-hoc solar construction monitoring is causing you delays

• Accurate surveying for project planning is vital to ensuring that solar plants reach their full potential. For example, the location and plan for PV installations will impact how much energy a plant produces during its lifetime. Planning that is based on high resolution terrain mapping, as well as shade analysis, will ensure maximized yield and long-term profitability. 

• Compliance with plans during construction is performed primarily by field engineers who manually compare drawings to make sure as-built is according to design plans. Because many engineers are supervising multiple construction projects, sometimes in remote rural locations, deviations from the design are often caught when it’s already too late to correct. 
• Monitoring  the progress  of construction means keeping track of the deadlines, budgets, plans, and regulations of projects. Since many construction managers are in charge of multiple projects (or one huge project) they often rely on third-party updates from contractors and subcontractors. These tend to be done using paper spreadsheets or online ones, and are difficult to keep track of when so many vital details need to be charted. 
• Enforcing warranties during commissioning kick starts only once solar panels are generating electricity for the first time, because these types of faults are not visible to the naked eye. In order to detect malfunctioning panels, a thermal inspection needs to take place, and the data needs to be analyzed to detect thermal anomalies. Only then can warranties be claimed, based on geospatial records of all anomalies detected. 
• Managing the masses of data that come from different sources, tools, and applications is a must for any company that wants to follow regulations, identify problems before they cause damage or failure, understand which parts of the project are on schedule and what areas are experiencing problems.

A single sophisticated software can streamline the entire process

Percepto AIM is a cloud-based software that manages all visual data captured by drones or any other visual sensor, on a digital twin of your site. It serves as a centralized repository of data for all your solar farms, from pre-construction to operations, and automates the transformation of the captured data into actionable information.

Working behind the scenes, but accessed from an intuitive dashboard, AIM helps optimize and streamline solar construction for groundwork, construction, commissioning, and operation and maintenance. 

• Engineering – AIM mapping tools generate 2D maps and 3D point clouds, mapping the terrain for site surveying and planning. To accurately estimate and monitor earthwork, AIM data analysis tools can produce detailed reports that measure slopes and pile volumes, as they change over time.
• Construction – AIM software provides visibility into what is happening on your site and tracks construction progress on a daily basis–even from remote locations. The software object counting algorithms don’t only count objects but also segment them according to type. It can share detailed updates on how many poles or panels are installed each day, and generate updates on progress vs. goals, all on a dedicated dashboard. In addition to detailed progress reports sent to all stakeholders on a daily basis, the construction progress is also displayed on a map timeline, where you can not only track progress in numbers but also in geolocation, as all data is displayed on a geospatial map of your site. 
• Commissioning – Once the electrical stage of construction is complete, the site is ready to start producing electricity. Percepto AIM data analysis tools produce detailed PV anomaly reports, accurate to the cell level, on a geospatial map of your site.  This makes it easy to streamline warranty enforcement and hand over a fully operational site. 

Meet increasing green energy demands with Percepto AIM 

For solar asset owners and EPC contractors, meeting the increasing demands for green energy poses several new challenges. And when scaling up construction to meet those demands, having one, easy-to-use software platform is key to staying on top of your projects.

The recently announced US nationwide waiver for Beyond Visual Line of Sight (BVLOS) operations means the skies have finally opened for expanded drone operations. The BVLOS waiver is good news for the entire drone industry, and great news for any company that needs to conduct remote visual inspections of industrial sites. 

But it’s not enough. While BVLOS approval opens new opportunities for drone-based visual inspections to take off (pardon the pun), it doesn’t solve the two major challenges of visual inspection: efficient, frequent data collection and visual data management. Even if organizations have advanced to using drones for visual inspection, many still struggle with one or both of those challenges because the technologies and processes they use only partially address their needs.

Challenge 1: Collecting high quality, consistent visual data daily

Effective industrial site visual inspection depends on the ability to collect the right data consistently and frequently in order to establish a baseline of all site assets’ health  and identify any deviations from that baseline. That requires:

• Data quality: Collecting high-quality RGB and thermal data requires drones that are equipped with advanced cameras and sensors. 
• Data consistency: Piloted drone flights are never identical. Different flight paths, differences in the image capture angles, different times of day…. Even slight variations make data analysis challenging.
• Data collection frequency: The frequency of piloted drone flights is limited by the high cost and low availability of drone pilots and the time it takes to manually upload and integrate siloed data.

A drone-in-a-box solution that is ruggedized and can be deployed on-site and operated year round,  with advanced payload and management software, solves the data collection challenge by enabling fully remote, high frequency, autonomous visual inspections. Inspections can be scheduled to run as frequently as needed (even multiple times a day) and every flight is identical.

That’s one of the two major challenges solved. The question then becomes: how do you get value from all that high-quality data?

Challenge 2: Leveraging high-quality big data 

Turning large amounts of data into insights comes with several challenges:

• Managing large data sets: Frequent, thorough inspections generate massive volumes of visual data. Sifting through it all to find the relevant images or videos for analysis is like finding a needle in a haystack. 
• Siloed data: In many cases, data is saved on shared drives or on dedicated softwares not accessible to all stakeholders. A drone pilot may collect great data, however that data in many cases is not getting to all relevant stakeholders, as the process of distributing it is manual. Same goes for insights gleaned from this data.
• Analysis: Analysis, by and large, is being done using the data collected in the most recent drone flight. It is not being compared with historical data, which means you don’t have an up-to-date, comprehensive understanding of defects and issues on your site.  

Many available market solutions do a great job handling specific steps in the drone inspection and monitoring workflow. But only a very few provide a comprehensive end-to-end solution that automates both data collection and data management. For instance, many drone solutions focus only on data collection and rely on third-party service providers to handle data processing and analytics. This creates a disconnect and lack of standardization that make it difficult for companies to fully leverage the potential of autonomous drone inspection and achieve real value. 

A single end-to-end solution for the entire data collection and management workflow is a must to fully automate visual inspections and leverage remote operations to enhance site reliability, safety and sustainability.

Industrial leaders are seeing the benefits of an end-to-end solution

Leading companies are already using end-to-end autonomous drone solutions to meet a variety of needs. In 2019, Delek US kicked off a plan to use autonomous drones with AI-based analytics for refinery surveillance, reconnaissance, and monitoring. Realizing that an end-to-end solution would be critical to achieving their “refinery of the future” goals, they looked for a solution with the right hardware and software to collect the data they needed and ensure that it would be organized, analyzed quickly, and delivered to the right decision makers. 

Delek implemented Percepto’s Autonomous Inspection & Monitoring (AIM) software for automated visual data management from upload to insight together with Percepto Air’s drone-in-a-box (DIB) equipped with an integrated OGI camera to detect emissions in real time and reduce product loss. The team is already seeing results on the ground and plans to significantly expand the implementation of Percepto solutions in the near future. 

Delek isn’t alone. Koch Fertilizer integrated an end-to-end inspection solution into their workflows at their Enid, Oklahoma plant. For the fertilizer giant, having automated data collection and visual data management is critical for their organization to meet its productivity and ESG goals. The organization was particularly interested in enhancing employee safety, by sending in automated robots to perform risky inspections. 

In a recent storm, for example, the solution enabled Koch Fertilizer to get their operations up and running quickly and safely, sending in autonomous drones to inspect critical infrastructure with Percepto AIM quickly analyzing their plant health status. For Koch, being able to do all that  without putting their employees in harm’s way is critical, enabling them to ensure plant reliability and safety.   

Autonomous drone inspection can also play a significant role in emergency situations like wildfires, where an on-site drone-in-a-box solution is the only feasible way to conduct monitoring. When the massive Big Hollow Fire shut down human access to a large area of Washington state in 2020, Verizon needed a way to make sure that fire, heat, water, or smoke damage wasn’t interrupting critical rescue and firefighting communications. That required powerful software that could enable remote operations and analyze Verizon’s infrastructure quickly to alert the team about any failures. 

With Percepto’s comprehensive solution, Verizon was able to respond quickly, dispatching autonomous drones to inspect and monitor network infrastructure from 1600 miles away. Percepto’s Autonomous Inspection and Monitoring (AIM) delivered AI-powered insights that allowed Verizon engineers to confirm the integrity and operability of its infrastructure without putting any personnel in harm’s way.

With a single end-to-end inspection and monitoring solution, these organizations and others like them can conduct more frequent, higher-quality inspections and detect and act on problems before they lead to downtime. Employee safety improves since inspections of hazardous areas can be conducted remotely. Stakeholders benefit from quality, data-based, actionable insights that deliver maximum value to their enterprise without pouring endless resources into visual data management. 

A complete solution is more than the sum of its parts

As those companies have learned, automating one component of the solution isn’t enough. A great drone-in-a-box solution creates big data challenges when it isn’t paired with advanced data management. A great data management and analysis platform can’t deliver the same value without frequent, high-quality, consistent data collected from an advanced drone-in-a-box.Bottom line – companies have much to gain from choosing a complete solution that automates everything. From data collection to data processing and analysis.

If you’re using drones once a year for PV thermal anomaly detection, that’s great. You’re running your site exactly at industry standards today for preventive maintenance. And if the benefits for deploying drones for solar panel inspections more frequently outweighed the costs, any logical site manager would be running inspections more frequently. So, do the benefits outweigh the costs? 

The answer isn’t always a big, all-caps YES (and that’s coming from us).

When considering adopting drones to help maintain PV farms, there is a delicate balance between productivity and OpEx costs. Optimizing productivity requires keeping a close eye on the entire solar array to spot issues like faulty trackers and panels obscured by dust or vegetation. And how clearly you can see the solar array depends on how you’re looking at it. And how often. But, reaching the right balance is a different analysis for each facility.