Researchers at two German universities have tested an experimental automatic landing system on a light aircraft that uses optical information from onboard cameras and satellite positioning data, instead of ground-based navigation aids.

The system was tested in May at Wiener Neustadt airfield in Austria on a modified Diamond DA42 piston-twin owned by the Technical University of Munich (TUM), which developed the technology in co-operation with the Technical University of Braunschweig under a project dubbed "C2Land", TUM says.

Large aircraft have been equipped with auto-land systems for decades, and follow radio signals from an airport's instrument landing system (ILS) during approach.

More recent navigation systems are capable of using satellite positioning data for landings at airports without an ILS. But as TUM notes, such procedures require ground-based augmentation systems to enhance the precision of satellite signals.

The vision-assisted auto-land system uses optical information from two cameras – one in the visual range and the other in the infrared range, for operations in low-visibility.

Image processing software developed by Braunschweig's team recognises a runway and determines the aircraft's position relative to it.

That information is then used by TUM's flight-control system for comparison with satellite data and to calculate a virtual approach path for landing, the Munich university says.

Video footage of the trial indicates that the system is capable of operating an automatic approach from altitude through to flaring and touching down on the runway, without manual control input by a pilot.

TUM says that the test aircraft was equipped with a fly-by-wire system – in addition to the DA42's conventional, mechanical flight control system – and featured an advanced autopilot developed by the university.

The autopilot's software was amended with several functions to facilitate automatic landings.

While autopilot systems are available for general aviation aircraft today – which typically have mechanical flight controls – TUM research associate Martin Kugler acknowledges that the existing vision-assisted auto-land system would not be suitable for installation.

Existing, off-the shelf autopilot systems for such aircraft tend to have less control authority than the fly-by-wire system on TUM's test aircraft, he says. But he foresees future demand for the technology as autopilot and flight-control systems on light aircraft become more sophisticated.

TUM also sees demand for its auto-land technology on unmanned air systems and, in particular, on electric vertical take-off and landing aircraft.