Using drones to inspect an industrial chimney

The era of using drones for inspections is well underway. Whether they are being used for an individual inspection, such as for freestanding water towers, or for larger, scalable projects, drones are becoming a standardized inspection tool.

Venezuela is yet another location where this use of drone technology is making a difference. Drone Pixel is a company based in Caracas that provide advanced RPAS inspections and applications. They’re a team of engineers using modern equipment and techniques to assist engineering projects. Now, they have integrated Pix4D and ESRI into their solutions to provide results for their clients. Drone Pixel's projects include site inspections equipment like telecom assets and power transmission lines as well as construction site analysis and insurance claim surveys.

Using drones saves Drone Pixel time and keeps their team safe on the ground

Industrial inspections with drones

In May 2022 Drone Pixel was contracted to assist on a project in the Aragua State area. Through their previous experience in using drones for vertical structures and industrial inspections, they have reduced risks by 99%, and cut work times by 60%. On top of that, they have cut costs by 30% through the time-saving and improved safety standards. Drone Pixel is seeing a significant return on investment for using drones for inspections. They knew they could use their knowledge for this project in Aragua.

The Hacienda Santa Teresa is a cultural heritage site that currently serves as a distillery and industrial production centre for the internationally renwoned St Teresa rum. At the factory, one of the chimneys is due for repairs and maintenance work. It dates back to the early 1900s and has a series of deteriorating elements that mean it could be a danger to production, or even at risk of collapse. Drone Pixel was asked to inspect the chimney and provide information and data that the owners could use to plan the chimney's restoration.

To do this, Drone Pixel chose to use PIX4Dscan and PIX4Dinspect, two specialized products designed for drone inspections. PIX4Dscan is a flight planning app with automated flight paths to assist inspections, capturing the fewest images possible whilst gaining complete coverage of the asset. PIX4Dinspect is a cloud-based inspection platform that uses machine learning and artificial intelligence to process and assist in asset analysis. The workflow between the products is seamless and the results can easily be shared in an automatically generated PDF report or via a sharing link, allowing for a fast turnaround of results.

The industrial chimney was over 60 meters tall, making it a challenge to manually climb and inspect

Project details

Inspecting infrastructure for restoration

The entire project from data collection to delivery took just 3 days, and had a high degree of accuracy with a ground sample distance of 5cm. The chimney is 62 meters high, and surrounded by other important infrastructure. It is not entirely straight, with some elements sticking out or mounted against the sides of it. The inspection needed to analyze the current state of the structure before identifying and quantifying any immediate risks. As it is a historical landmark, Drone Pixel’s client BRS Ingenieros needed as much information as possible to make informed decisions about the potentially complicated restoration project. The final outputs would be shared with financial, engineering, and strategic planning teams at BRS Ingenieros, emphasizing why PIX4Dinspect (with its high shareability) was ideal for this project.

Drone Pixel found the data collection and processing workflow very intuitive. Thanks to the automated flight plans in PIX4Dscan, the photos were captured without the need for manual control. No people were put at risk by having to climb the chimney and take photos because the drone captured the entire structure with a high degree of detail.

The drone camera picked up cracks in the concrete as well as the underlying metal and rust

PIX4Dinspect processed the images and created an intelligent digital twin enhanced with machine learning and artifical intelligence. The twin was analyzed by the Drone Pixel team, who were particularly impressed by the automatic rust detection feature, which was highly accurate. Before sharing their results with the client, Drone Pixel shared the digital twin internally to make sure all the relevant engineers had approved the results, before they shared it externally. The rust detection helped identify fractures in the structure as well as the loss of coating on metal elements. The concrete falling away was also identified and annotated within the software. Annotations on the digital twin will inform construction teams about these faults and help them keep safe during restoration in case the old conrete falls or gives way.

With PIX4Dinspect, Drone Pixel identified problems and could inspect them with the image view to see more detail

Delivering a tailored inspection report

The rapid turnaround of this project highlights how drone inspections can empower significant time savings in inspection projects. Drone Pixel were pleased to provide their client with a highly detailed PDF report that was automatically generated by PIX4Dinspect as well as the 3D model, which their client can use to see even more detail and add their own annotations or notes.

The final 3D model was accurately measured and included some of the terrestrial area to assist restoration work

Thanks to using PIX4Dscan and PIX4Dinspect, Drone Pixel reduced their working time, increased the operational efficiency, and produced an AI-enhanced digital twin. The model and report together provided all the details needed for this project and complete the inspection process.

Drone Pixel believe that the success of this project, with such an influential client in Venezuela, will help them grow their business operations. The proven cost savings and efficiency improvements cannot be overstated as it has revolutionized how Drone Pixel operate in the 5 years since they started using Pix4D solutions. They are focused on getting the best results as fast as possible without compromising on the quality of their outputs, which is possible with Pix4D.