It’s been a smooth flight, and you’ve spent it engrossed in a juicy novel on your tablet. You barely noticed the captain’s announcement about a half-hour ago that “we’ll be starting our descent into Vancouver in a few minutes,” and that “there are low clouds around the airport.”
Next thing you know, the plane thumps onto the runway and begins to slow down. Looking out the window, about all you can see is the glow of the white, runway edge lights zipping past, shrouded in thick cloud.
Just how did the pilots safely land in this pea-soup?
They had high-tech help.
Based on the pilot’s inputs, an autopilot calculates a path though the air, and it’ll move an aircraft’s control surfaces to fly straight and level, through turns, climbs and descents, or to follow a specific flight plan.
The control surfaces are moveable, aerodynamic devices on the plane’s wings and tail that allow the pilot to control the flight attitude (the orientation of the plane in relation to the horizon).
The first aircraft autopilots were developed more than a century ago by American Lawrence Sperry, which he demonstrated in France in 1914.
In 1937, US Army Air Corps experiments in automatic landing systems resulted in the first successful autopilot-controlled landing.
But the technology didn’t mature until the 1960s when British European Airways (BEA) – an ancestor of British Airways – began flying automatic landings using the Hawker Siddeley HS-121 Trident short (and later medium-range) passenger jet.
Driven by the often bad weather conditions across the UK and Europe, the Trident’s ability to perform what was described as a “blind landing” allowed BEA to maintain its flight schedules.
Fifty years later, virtually every modern airliner from the smallest regional plane to the largest wide-body jet has autoland capability.
Follow the beams
When clouds surround an airport, pilots have been able to find the path to the runway for decades by using an Instrument Landing System, or ILS.
Ground-based transmitters project one radio beam straight down the middle of the runway, and another angled up from the runway threshold at a gentle three degrees.
During an approach, a display on the aircraft’s panel shows pilots whether the plane is to the left or right of the runway – on the localizer beam – and above or below the descent path, called the glideslope.
To complete a safe landing, pilots must be able to see the runway at a specific decision height (DH) above the ground. A minimum horizontal visibility is also specified, called the Runway Visual Range, or RVR. The RVR is measured on the ground, and the information given to the pilots before they begin an approach.
At DH, if the pilots can’t see the runway environment – which can include the runway’s high-intensity approach lights – the crew must climb away from the airport, and either try again or go to an airport with better weather.
Generally, the minimum altitude on an approach is at least 200 feet above the ground, depending on the airport’s location and surrounding terrain. With additional equipment on board and a tightly calibrated ILS system, minimums can drop to 100 feet.
ILS approaches are designated by category. CAT I has the highest minimums, CAT II is lower, and three levels of CAT III approaches drop the decision height to zero altitude and zero visibility – a true autoland and rollout.
The first aircraft certified to fly a zero-zero approach was the Lockheed L-1011 TriStar, in the 1970s. That early wide-body was equipped with a Collins Aerospace autopilot, and the company continues to make sophisticated, advanced-tech systems for many airliners, including the Boeing 737MAX and upcoming 777X, and the Airbus A220.
But it isn’t just about computers, explains Craig Peterson, senior director of commercial systems marketing for Collins Aerospace.
“The flight crew involvement is key in the entire autoland operation. Not only does the airport facility and the aircraft need to be approved for autoland operations, the flight crew also needs to be trained and be approved to perform them. The autoland system is a high-integrity complex system due to the safety-critical nature, and the flight crew has to continually monitor the system operation and performance.”
On a CAT III autoland approach, the pilots cross-check the plane’s speed and path to the runway and are ready to take over from the computers, should a system fault occur.
To become highly proficient and certified to fly autoland operations, pilots endure intensive and sweat-inducing training sessions in ground-based simulators – ones with ultra-realistic graphics displays, controls and cockpit movements that accurately duplicate the flight experience.
“The simulator is so accurate, and we can reproduce the worst weather conditions,” explains John Mulder, chief pilot, flight technical operations for Calgary-based WestJet. “It’s the best place to train for CAT III approaches. In the simulator, we tend to go right to the limits and the capabilities of the aircraft.”
The Canadian airline has a large fleet of Boeing 737s, including the newest 737MAX, and will soon be flying the globe-spanning Boeing 787 Dreamliner.
WestJet’s 737s are equipped to fly CAT III approaches down to a decision height of just 50 feet, the altitude at which the captain must quickly decide whether to let the plane land or to abort the approach.
“There’s a radar on the bottom of the aircraft that’s measuring our distance to the ground. We’ll be past the high-intensity runway lights, and at that point we’re going to be over or past the runway threshold,” said Mulder. “We have to have the landing environment in sight or we’ll go around.”
Many airliners can fly full CAT III zero-zero approaches, with additional on-board computers and systems that automatically track the runway centerline after touchdown, apply the brakes, and bring the plane to a safe taxi speed.
Firm but fair
Pilots know that passengers will give high marks for a smooth landing, when it feels like the wheels have simply rolled onto the runway.
But in rainy or snowy weather, that isn’t necessarily the best way to put the plane down, according to Mulder. “We actually like to have a landing that’s a little bit firmer to break the wheels through any kind of contaminants, to get them firmly on the ground so that braking can be effective.”
That kind of touchdown is programmed into the autoland systems, even though passengers might feel like they’ve landed on an aircraft carrier instead of a runway.
So, who makes the more consistent landings, the pilot or the computer?
“Well, to be honest, the firm landing is the better landing,” said Mulder.
“I guess I’d have to give that to the autoland system.”