If I interpret Denti’s operating limits correctly (#39), then LAND 2 may be part of a fail operational system (hybrid) enabling either continued autoland or reversion to manual landing with HUD. I would be surprised to learn if pure manual landings were authorised in RVRs below 200m for system failures below DH; a GA is normally mandated after any failure. There are many interpretations of operational clearances (including interpretations made by the regulators) – not all are valid.
Where a ‘super fail passive’ system enables an operation in RVRs below 200m, normally a GA is the mandated (regulator) procedure below DH, vs manual landing above 200m. I would tend to agree with Clandestino; not on the safety aspects of ability, but of having one consistent operating procedure for all RVRs.
However, my experience with these systems is that for the very few (and rare) failures below DH in RVRs less than 200m, the vast majority of the crews continued with a manual landing (against SOPs). The reasons given were very low altitude and judgement of the safest option – such is human performance.
Also, the majority of the failures involved ‘inadvertent’ autopilot disconnects caused by the crew’s tight grip on the controls. Other failures were due to ILS transmission failure or interruption. Thus there is need for appropriate crew training and adequate guarding of the autopilot disconnect button – a design issue.
Re the crosswind limits, these are usually hard limits published in the AFM (vice the more normal advisory max demonstrated) and result from the autoland certification. What many operators (and operational regulators) overlook is that humans have not been ‘certificated’ for manual landing in high crosswinds on a wet runway in low visibility.
Manual landing in 25 kts wet can be a demanding task in clear visibility, in fog it’s distinctly challenging. This area identifies a significant weakness in the regulations which require good crew judgement to apply lesser limits to their operation. Another good example is in blowing snow where there is a visual disorienting effect in additional to poor runway friction – laterally and longitudinal.
Autoland systems do not consider these aspects; it’s up to the crew, many of whom in my experience forget to attend to the ‘side’ issues of low vis operations when using automation.
There may be similar problems with HUD operations. The research quoted earlier evaluated development versions of HUD and noted “problems with division of attention” between symbology and the outside world, and “problems of presenting lateral guidance”, which perhaps indicate that HUD crosswind operations might have a lower margin of safety than normal operations – again good crew judgement is required.