ELUDING CRASHES IN CONFINED SPACE DRONE INSPECTION
ELUDING CRASHES IN CONFINED SPACE DRONE INSPECTION
Prominent tech investor Cathy Wood cites the 5 new “transformative technologies” gaining traction in the last decade and fostering innovation at a blistering pace. They are genomic medicine, AI, blockchain, robotics, and energy storage. The drone industry involves 3 of these 5 transformative technologies. Innovations in robotics, AI, and energy storage (batteries) have led to dramatic improvements in UAV technology in the industry’s short life span. However, like many nascent industries, sometimes the hype outpaces the reality as companies seek headlines and “first-mover status” – the drone space is no exception.
THE FIRST APPROACH
One essential element of confined space drone inspections is the ability to safely navigate tight spaces in a GPS-denied environment. The first approach was to construct a cage that surrounds the drone, mainly to protect the propellers from a collision that could render the drone inoperable. Cage designs varied, including geodesic dome patterns, vertical ribs, and checkerboard patterns. All of these designs provided the ability to literally bump into walls or objects while conducting inspections, allowing the camera to get within 6 inches of its target for close-up, high-detailed data-capture.
WORK SMARTER, NOT HARDER
In the last few years, others have claimed the ability to eliminate the possibility of a collision altogether through the use of a combination of sensors and software. The most common type of system being used today is LIDAR (Light Detection And Ranging). In layman’s terms, you can think of LIDAR as similar to radar but using pulsed light instead of microwaves to detect objects in the drone’s proximity.
THE CSDI PERSPECTIVE
While we predict that this will become the standard approach over time, CSDI Group remains skeptical (for now) that obstacle avoidance alone offers the best results for commercial confined space inspections. The systems being used today have limitations in detecting objects thinner than 1/2 inch in diameter, making risk-free flight unattainable through using LIDAR and software alone. Additionally, the buffer that is set for obstacle avoidance requires the drone to remain a minimum distance from any object. This buffer can often be manipulated, though it is not recommended. In confined space inspection, gathering high-detail data is a priority; limiting the distance between your drone and the asset can result in a lower-quality inspection. Other complications have been illustrated in high-wind environments, where the drone software does not correct for wind displacement quickly enough to avoid a collision.
The pace of innovation in UAV technology is impressive. One day, fully autonomous platforms that can navigate and inspect assets without human assistance will be the norm. But, in our opinion, the technology is not there yet. The loss of a drone in a permit-required confined space is a failed mission. Until there are significant improvements in LIDAR systems and their accompanying software, we prefer to inspect assets with a caged drone.