Unmanned Systems Technology 004 | Delair-Tech DT18 | Autopilots | Rotron RT600 | Unmanned surface vehicles | AMRC | Motion control | Batteries

38 avionics-qualified components are tested extensively to show much higher reliability figures than commercial components. System designers are also increasingly including degradation detection algorithms in the autopilot, which identify failure modes in real time, identify their cause and respond appropriately. This also gives the opportunity for predictive maintenance, allowing boards and subsystems to be identified as beginning to fail and replaced before they do fail. As the autopilot takes in the sensor data from around the airframe, this is a natural place to implement these predictive algorithms. Software development The third safety factor is the software development process. Avionics industry standards such as DO-178C define a process for developing, testing and qualifying the code that runs on the autopilot to ensure its safe operation. The leading RTOSs are certified to be used with this process, mainly for commercial airliners and military manned aircraft, but it also reassures regulators that the software for an unmanned system has been developed with safety in mind. DO-178C is not the only way of demonstrating safe software. For example, one major autopilot supplier uses its parent company’s quality systems for developing commercial and defence aircraft subcomponents. The fourth factor is regulatory. The levels of required safety are clearly defined for airliners and other aircraft, but not for unmanned systems, and this is part of an ongoing process that is having an impact on the development of autopilot software. The Atlante RPAS (Remotely Piloted Aircraft System) UAS, for example, uses a commercial RTOS for its autopilot, developed under a contract for the Spanish government. In October 2014 it was the first large unmanned craft to be submitted by a European company to the European Aviation Safety Agency for civil Type Certification. This marks an important milestone in opening up the European civil aviation market for RPAS operations in the 150-plus kg Maximum Take-off Mass category, and it is also a key part of developing a European airworthiness regulation for the RPAS to open the aviation market to the civil use of remotely piloted aircraft systems in a safe and sustainable manner. There are also regulatory issues that affect the function of the autopilot. For example, safety requirements may be addressed by defining air corridors for the use of unmanned systems, and this is currently happening in the US for airborne delivery systems. This will have an impact on the programming of autopilots as developers will have to demonstrate to regulators that the craft cannot fly outside these corridors. This can be done by ’geo-fencing’, linking data from the satellite navigation system to the autopilot which has the corridor boundaries programmed in. This then constrains the autopilot to operating only within the corridor, but it means the autopilot has to have this capability built in. A fifth approach to safety that is gaining acceptance is distributed computing. Autumn 2015 | Unmanned Systems Technology Triple redundancy can be a more costly approach but provides greater reassurance of safe operation in all circumstances (Courtesy of Micropilot) The Atlante was the first UAS from a European company to go through European airworthiness certification, and uses an autopilot based on the VxWorks real-time operating system (Courtesy of Wind River)