Robot-assisted ship inspections sail towards certification
Ship inspection surveys are essential to ensuring the safety of a vessel throughout its life. Conducted by classification societies, these periodic surveys include extensive examinations to verify that a ship’s structure, machinery and key systems remain in a condition that satisfies all rules and regulations. Traditionally, these surveys are conducted by highly qualified personnel. However, as ships become both larger and more complex, many surveyors are using technology to reduce the risks and costs associated with inspecting hazardous, harsh and dirty environments. “Wide volumes with significant heights, like a bulk carrier’s cargo hold, require costly means of access such as scaffolding or hydraulic platforms, and pose a significant risk to the surveyor,” says Alessandro Maccari, marine research and development director at RINA Services. “On the other end of the spectrum you have narrow and confined spaces, which pose hazards related to access, mobility, ventilation and cleanliness.” Robotics and autonomous systems (RAS), such as drones and crawlers, are becoming an attractive alternative to traditional ship surveys. Not only are they more affordable and safer, RAS-based inspections can acquire and process visual data faster without compromising its accuracy, greatly reducing the time vessels are docked in harbour for inspection. However, before robots become a standard tool in the surveyor’s belt, they must first be certified. The certification process requires testing, metrics, protocols and guidelines – which is exactly what the ROBINS (Robotics Technology for Inspection of Ships) project set out to do.
Traditional vs RAS-assisted inspections
The goal of ROBINS was to establish a framework for assessing the equivalence between traditional and RAS-assisted inspections, defining measurable qualitative and quantitative standards. “Our aim was to fill in the existing technology and regulatory gaps for the adoption of RAS in the life-cycle surveys of ships,” adds Alessandro Giulio Grasso, a project coordinator at RINA. To do this, the project improved and tested a variety of RAS. These included a collision-tolerant drone for inspecting irregular confined spaces like ballast tanks, along with a semi-autonomous drone for surveying bulkheads and structures inside large cargo holds. ROBINS also developed a crawler: a small, agile robotic vehicle for close-up surveys that can climb steps, manoeuvre around corners, and reach and probe structures. Furthermore, the project developed a range of innovative software tools for autonomous RAS inspections. “These programmes leveraged such emerging technologies as LiDAR, photogrammetry, machine learning, artificial intelligence, and 3D model augmentation for navigation, localisation and data acquisition,” notes Maccari, who helped coordinate the project.
Faster, cheaper and safer
Next, test campaigns were completed in field trials and in testing facilities. These were used to measure the RAS’ capabilities, compare them to traditional surveys and, based on this, develop performance standards. “What we found was that these systems gather information faster, cheaper and safer,” notes Grasso. “Not only are these technologies able to detect and accurately map defects throughout a ship, they also are highly cost-effective in hard-to-reach areas.” Other benefits include the ability to use RAS in most ship types and offshore units and to provide instant feedback from vessel to shore. Drones also have the advantage of being easily deployed and operated by a single person without any need for extensive – and costly – safety equipment. “By proving equivalence between RAS and traditional inspection activities, we’ve opened the door to the massive adoption of RAS in Class and Statutory ship inspections,” concludes Maccari. “Not only will this create safety and cost benefits for ship operators and surveyors, it will also stimulate the EU robotics industry, unleashing the economic potential of new markets.”
Keywords
ROBINS, robot, ship inspections, robotic and autonomous systems, drones, crawlers, surveyor, cargo
