The origins of the autoland: not foggy at all By Captain Robin Evans, Senior Log Contributor unbuckled my safety belt, dived over the side of the fuselage, and after two or three seconds of fall, pulled the rip cord. So said Charles Lindbergh of the final option for 1920s airmail pilots encountering fog. If airfields were obscured, pilots would light parachute flares and spiral down around them. Forced landing elsewhere would mean entraining the mail at the nearest railway station: a commercial disruption to a fledgling air link. In his year of airmail tenure before making transatlantic history in 1927, Lindbergh became a parachutist twice, once hearing his aircraft continue to circle around him as he descended in darkness. This early solution is unthinkable by modern standards; as jeopardy and expense grew, we needed something less disposable Guidance Early references to low visibility at any altitude used the generic fogs; IMC was not yet fully understood. Pilots first needed a systematic way of landing at all. Early methods suggested rockets or balloons perhaps containing an observer suspended on a nominal path. Initial approach-path lighting appeared in the 1930s, and was refined into the approach-path system of the 1940s that began to indicate perspective and displacement. The first system we would recognise as a predecessor of the instrument landing system (ILS) is German: the 1930s Lorenz beam. Dots and dashes were transmitted by ground array, the two overlapping with a constant audio tone when the aircraft was centralised, later indicated on a gauge. US Army captains Carl Crane and George Holloman made what is considered the first automated landing in August 1937. After two years of testing, Holloman overflew four sequential beacons, devised by Crane. Established laterally, the final beacon commanded a powered descent. Flown into the ground, a squat switch commanded idle thrust and braking. Wartime required the safe return of assets during a European winter. With both sides racing for supremacy in the Battle of the Beams, could high-altitude principles be applied to low? One system, used by the Normandy Pathfinders, relied on signal timing between an airborne Rebecca (for recognition of beacons) transceiver and a Eureka ground transponder. This evolved into the beam approach beacon system (BABS), using call-response, transmitting lateral displacement by dots one side and dashes the other from a fixed antenna or Hillman van. Motivated by fears of limits imposed on their bombing by European fogs, the United States Air Force brought their own setup. In development since the 1930s, the Signal Corps System 51 (SCS-51) is the early forebear of our ILS. Shared with the boffins of the Royal Aircraft Establishment (RAE), this produced the first coupled approach, flown straight into the runway in 1945. In the interim, an alternative wartime solution existed: a network of gasoline burners plumbed alongside the runway. Developed by Dr John Main-Smith, of the Farnborough chemistry department, fog investigation dispersal ops (FIDO) was used operationally from 1943. Installed at more than a dozen bomber airfields, it prevented the loss of an estimated 2,500 aircraft and 10,000 aircrew. Earliest steps Various elements now existed, but no system combined their strengths. Represented by the RAE, Britain began a fundamental role. The Blind Landing Experimental Unit (BLEU) was founded at Martlesham Heath, Suffolk, in 1946, for the development of military and civil blind landings. The value of this work would be emphasised by the growing commercial focus of the jet age, particularly in Europe, where restrictive weather was most common. The BLEU initially defined a localiser with buried leader cables. An equivalent marine system was installed in New Yorks Ambrose Channel in the 1920s: an electromagnetic field created by miles of seabed cable. By switching between induction coils hung overboard, shipping could gauge signal strength and maintain a defined track. Obsolete by the 1930s, the Ambrose cable was lifted and used in initial American autoland research. Trials concluded that cable setups were expensive and impractical, but their use continued until wireless localiser resolution was sufficient. The offset location of the glidepath antenna required an altimetry cross-check: the radio altimeter. Its mechanical ancestor was the ground proximeter, developed to compensate for the lack of visual cues by night on military landing grounds. Dating to 1916 in the Royal Naval Air Service, a weighted cord was lowered from a drum beneath the aircraft. As it struck the ground, the reduced cord tension signalled an indicator lamp, commanding the pilot to flare. A variation occurred several years later, the lamp becoming the trigger to release pressure against nose-up trim. Deprived of radio aids in World War II, Ernest K Ganns Fate is the Hunter details a similar improvisation into Reykjavk a blind letdown by descending until the radio antenna cable snags the sea at 40ft. Work also focused on the workload and human factors of transitioning from internal to external cues. The relationship between visual perspective, runway visual range, descent rate and pilot response time at various speeds was tested extensively. Agreed with ICAO in 1949 was the Calvert array, a lighting rig replicating runway perspective, designed by E S Calvert, a research scientist at RAE Farnborough. Varsity WF417 and friends on the apron Flt Lt Pinkie Stark displays his look no hands routine on approach in a Varsity Collaboration American work on SCS-51 became the basis of the ILS, accepted by ICAO in 1948. Beam resolution was defined by a given deviation being sensed and indicated, then perceived and corrected at a nominal rate of descent in the final stages of an approach. This began to inform Cat 1 limits. The accepted date for the first demonstration of an entire system, including autothrust, is 3 July 1950. More than two decades of work remained. Despite pilot resistance and American pressure for synthetic (headup display) runway indications, the BLEU conclusion was that a fully automatic system was needed, not one reliant upon pilot response. This meant coupling powered controls to a feedback loop, which was accelerated by automation developments and industry partnerships. Many aspects had to be resolved: announcing component failure, extending approach lighting into the touchdown zone, and de-crabbing. The BLEU began testing using 1950s military workhorses the Anson, Devon and Varsity. The official requirement for an all-weather landing capability on the Vulcan arose in 1954, motivated by returning warheads home safely. In 1957, the BLEU moved to RAE Bedford, the sister site to Farnborough and a crucible of wider flight research. Imagine a British version of the Lockheed Martin skunk works, fuelled by tea and crumpets. Many projects of relevance today took place here: wind-tunnel modelling; swept-wing geometry; the PAPI (up to 15o); fly-by-wire; and aquaplaning after the 1958 Munich disaster. An observation from wake-turbulence trials of a later decade reveals the empirical nature of testing. A DH125 was vectored close behind a Concorde. A muffled and multi-voiced reply indicated some flight-deck difficulty eventually, the pilot said that was a bit too close, so much so that some of the flight-deck roof lining had come adrift. For every celebrated test pilot of the post-war heyday there were unsung counterparts focused on work equally fundamental to modern standards. One was Flt Lt Laurence Stark, nicknamed Pinkie on account of his rosy cheeks from fetching coal for the mess during wartime winters. A Typhoon ace awarded the Distinguished Flying Cross and bar, he was shot down over France, evaded capture and later spent seven years with the BLEU. Initially, pilots maintained roll and yaw control into the touchdown zone with the flare starting automatically. Eventually, the system maintained the localiser throughout. Because of the decreasing width of the glideslope beam, the vertical mode would revert to attitude freeze at a nominated radio altitude, typically 120ft. Any wind variation thereafter was deemed of negligible effect. At 50ft the flare computer took over the pitch channel, assessing radio height against its rate of change. Thrust was reduced and the yaw channel returned to the crew. This look no hands routine was demonstrated to the Duke of Edinburgh in a Canberra in 1959. BLEU superintendent John Charnley reported feeling distinctly nervous about this, because the Canberra had only a single channel military setup. Though apparently impressed, there is no record of what HRH said. From left, Captains Carl Crane, George Holloman and assistant Raymond Stout with their autolanding Fokker Imagine a British version of the Lockheed Martin skunk works, fuelled by tea and crumpets Power of three After five years work, the Vulcan became the first quad-jet to autoland, on 22 December 1959, its single-channel system certified in 1961. Authorities had specified an accident rate attributed to the autoland system of no more than 1 per 120,000 landings (military) and 1 per 10 million (civil). The latter case, therefore, required redesign as a multiplex system either dual (the VC-10 employed a constant-evaluation loop and auto-changeover) or triplex. The Trident was the first jet airliner to be designed from scratch to autoland, with triple circuits throughout: engines, hydraulics and autopilots (and three pilots in the flight deck!). As its avionics required more space beneath the flight deck, the nose gear was offset and retracted sideways, a rare design feature. British European Airways (BEA) was particularly keen on the potential of autolanding, given the prevalence of foggy winters, exacerbated by coal-fired power stations, in the western European network. Londons notorious pollution-fog pea soupers had already prompted the Clean Air Act of 1956. A Trident made the first autoland by a production airliner at Bedford in 1963. Trident G-ARPR pioneered it commercially on 10 June 1965 between Le Bourget and Heathrow. Captain Eric Poole a graduate of the Battle of Britain, who joined BEA shortly afterwards landed himself in flight history by not physically landing himself. This was announced afterwards to the passengers, who were presented with souvenirs. Decades of work were distilled into the public message look no hands, a slogan reinforced by PR footage of a pilot explaining the system to camera all the way to touchdown while looking rearwards. Operational testing continued in passenger service, the Trident programme amassing 40,000 autolands in 20 years of development to prove Cat 3C certification (zero-zero, though not required in practice today). This contributed to the formal definition of autoland minima and the monitored approach pilots monitoring the machine, as well as each other. The BLEU became the Flight Systems Department in 1974. With the thousands of autolands performed at Bedford by day, night and in all weathers, there were no accidents. The autoland programme caused no damage to any aircraft, through good fortune, rapid and effective pilot intervention, but mainly through the cautious, step-by-step approach implemented at all times, reported the Bedford Aeronautical Heritage Group. The BLEU was a world-leader in the design of autoland systems, a status borne out when the FAA [Federal Aviation Administration] sent its research DC-7 to Bedford in 1961 for collaborative research. Full circle Much work carried out at RAE Bedford remained secret, and the airfield closed in 1994. It is now known as Thurleigh Business Park. In the meantime, a century of progress has been distilled into the simple passenger belief that autoland occurs at the touch of a button. That misconception is now reality, however: in 2020, Garmin released a GPSdriven emergency autoland function as part of its G3000 upgrade. With thanks to the Bedford Aeronautical Heritage Group (BAHG) www.bahg.org.uk This piece was inspired by a company-approved media request for technical assistance. Though beyond the scope of that project, the details uncovered were surprising.