On-aircraft inspections are often required as part of the maintenance of aircraft. Alongside this, in-service aircraft structures, such as fuselage or wing skins, experience damage or degradation during their lifetimes.
In the commercial aircraft world, most of this inspection is done manually, with handheld probes connected to non-destructive inspection (NDI) instruments. Manpower and cycle times can be high, while inspection repeatability, reliability and record keeping is generally low. Also, access requirements for difficult and dangerous locations can lead to human injury or aircraft damage.
NDI scanning systems are currently used for inspecting some aircraft. These systems generally require human access to the structure to align, attach, move, re-attach and detach. Correlation of the scan data to aircraft structure can be challenging.
What is MANDI?
Mobile automated NDI (MANDI) is a new area of inspection that uses a mobile (crawling, flying, or swimming) robot to reduce the requirements for human interaction with the structure under inspection. MANDI can be applied to the exterior or interior of structures in the factory, depot, or field.
MANDI is different from traditional automated NDI that uses multi-axis scanning bridges, and even standing pedestal robots. Its main advantages over these systems are that systems are portable, have a small footprint, and are much lower in cost.
Of course, crawling robots of one form or another have been around for some time. They are used for surveillance, bomb assessment and disposal, access to unsafe or limited entry areas, and even window washing. Specific NDI applications for crawling robots include storage tank and pipe inspection.
What makes the system work?
The requirements for carrying out in-service automated NDI for
aerospace structures are quite different from those encountered during manufacturing or post-production. Unlike the large monument NDI systems, MANDI systems need to be as low cost as possible, since maintenance budgets are always a concern for airlines and MROs.
They need to be relatively easy to use and not require an advanced degree to operate. Minimal human interaction with the structure is preferred, and labor should, in principle, be traded-off for the speed and consistency of automation. The robot must be safe for the operator and other personnel in the area.
Care must be taken that the robot does not damage the surface it crawls on, while being able to maintain attachment, position and direction during inspection. Automated guidance or feedback control capability is preferred, as it removes the operator from the challenges and errors related to direct human guidance of a robot.
One of the key features for MANDI is that it should allow easy registration of the NDI data to the aircraft. Obviously, identifying the specific locations of damage indications is important to their proper disposition.
Boeing has just completed a MANDI prototype system for aircraft inspection dubbed the Rover, which stands for ‘remotely operated vacuum-enabled robot’. The Rover is a motorized, low-cost, lightweight robotic crawler that can carry various NDI sensors (UT, EC, etc) for the inspection of aircraft structures. It is extremely innovative in its design and represents the integration of a variety of advanced features. Several of its important elements are off-board, including path planning, navigation, and the control of the NDE data collection system. By putting these features off-board the crawler, the size and weight of the Rover can be minimized, making it more nimble and increasing its maximum payload. Attachment of the crawler to the inspection surface is accomplished using a set of onboard vacuum pumps that create a suction though a series of floating cups.
Simultaneous rotation and translation crawling motion (called holonomic motion) is made possible for the Rover using four independently powered and controlled Mecanum wheels.
The Rover does not require a perfectly smooth surface; it is actually capable of traversing lap joints and raised or button-head fasteners without losing its attachment. It has an automated fall protection system to prevent damage to the crawler or injury to personnel working nearby. It also has joystick control for pre-scan positioning and post-scan removal of the crawler, and an optional onboard camera system for situational awareness, should the operator require it.
The Rover feedback control and NDE data registration is accomplished with one of two different methods: the local positioning system (LPS), or motion capture (MoCap).
The LPS is a Boeing-developed coordinate measurement device (with a motorized pan-tilt head, laser range meter, and video camera), with custom software to convert measured positions into the coordinate system of the target object. The system can be controlled remotely over the internet. The device can also be instructed to move the laser pointer to user-specified coordinates on the target. The LPS uses what is called an integration visualization tool (IVT), developed for 3D visualization on Boeing commercial airplane programs. The IVT allows display and manipulation of large amounts of CAD data for design reviews and analysis tasks. After an initial positional calibration, the LPS guides the Rover between points and checks, and adjusts its position at the end of each run. Several sets of encoder wheels track the Rover position temporarily between LPS checks. While accuracy depends upon the distance between the LPS and Rover, it is typically ±3mm at about 10m, which is generally sufficient for NDI scanning using an array. A slight overlap between scans is programmed in if an areas scan is required.
The MoCap system
The MoCap system can be used when greater positional accuracy is required for the Rover, or when multiple Rovers are running at the same time. MoCap uses off-the-shelf motion capture hardware initially developed for the movie industry, along with patented closed-loop feedback control technology developed by Boeing that leverages motion capture hardware for vehicle tracking and control.
This system consists of multiple stationary cameras with integrated illuminators on portable stands, or fixed-position mounts placed around the target object. This system tracks unique patterns of retro-reflective markers placed on the crawler to determine its position and orientation. The system is less portable than the LPS, and more costly, so its use may be more limited to production inspection.
The Rover was completed in December 2013 and a demonstration of the system was performed on a 737 test fuselage (see figure top right). Both the LPS (with encoders) and MoCap guidance methods were demonstrated, with the Rover independently following a pre-planned path. Visual and eddy current real-time data was collected, as the Rover stayed attached and crawled with holonomic motion around to the fuselage. A mobile boom, with a safety tether and power/communications cables, followed the Rover automatically as its base moved parallel to the fuselage, driven by its own Mecanum wheels.
MANDI is a new approach to NDI, where robotic crawlers replace current methods of sensor placement and scanning. They are able to reduce or eliminate the human interaction in the data collection process, and eventually, the analysis process as well.
MANDI has the potential to reduce labor costs and inspection times, provide improved data quality, and support ongoing maintenance management. It also ensures a safer working environment for inspection personnel and reduces the potential for damage to the aircraft during inspection.
Boeing’s Rover is a breakthrough MANDI system aimed specifically at the in-service inspection of aircraft structure. The development has involved the advancement and/or integration of multiple technologies into a system that has worked surprisingly well. The Rover is the first MANDI system to integrate off-board navigation, direct correspondence to the location on structure, an automated mobile safety/tether boom and onboard vacuum attachment. Future Boeing efforts will be focused on refining the Rover and identifying new applications for its use in the broader aerospace community.
Gary Georgeson is a technical fellow for Boeing Research and Technology in non-destructive evaluation (NDE) based in the USA.