Sometime between 1947 and 1956 there was an air show at Bolling Field. I was a kid, attending that show with my World War II pilot father. They were flying an empty B-17, remote controlled by a radio link from a second B-17 flying behind it, attempting to land the pilotless plane on the runway in front of the crowd. The remote-controlled plane hit hard. I can still see this in slow motion in my mind: The plane was sliding off the runway toward the reviewing stands where we were sitting. It was a pretty dangerous thing to attempt at a public show in front of a crowd. Does anyone have a memory of this?

Paul Friday, Washington

The B-17 that landed hard that day — Sept. 18, 1948 — was not pilotless, but it wouldn’t be quite right to say that Capt. Horace L. Spencer was at the controls. Well, he was and he wasn’t.

Spencer was what was known as the B-17’s “safety” pilot, seated in the cockpit and ready to take the yoke in case something went wrong. Piloting the drone in flight from a nearby B-17 “mother ship” was Lt. V.F. Grissom Jr. On the ground at Bolling was another pilot, Lt. Allan H. Hoover, who was to control the landing.

Spencer and Grissom had taken off that morning from Eglin Field in Florida, destined for Washington and an air show marking Air Force Day. The planes were showcasing an experimental technology introduced in World War II. During the war, pilotless aircraft served as missiles, delivering a punch to hardened, well-defended targets.

The idea was to strip old bombers — the term the brass used to describe these battered workhorses was “war-weary” — of all unnecessary equipment and pack them with explosives. Human crews would take off from airfields in England and bail out 10 minutes later after arming the charges. A pilot on the mother ship would take control, using radio signals to direct the drone and crash it into its target. The Army Air Forces called its effort Aphrodite, the U.S. Navy, Anvil.

The success rate was slim and the risks involved were enormous. A project Anvil mission cost the life of Lt. Joseph Kennedy Jr., older brother of the future president. On Aug. 12, 1944, the Navy flier and his flight engineer, Lt. Wilford J. Willy, died when their drone bomber — loaded with 21,000 pounds of explosives — disappeared in a fireball over England before they’d had a chance to bail out.

After the war, radio-controlled bombers were flown through the mushroom clouds of atomic bombs detonated in the South Pacific. The planes could gather data without endangering a human crew. It was one of these “atomic tested” B-17s that was scheduled to land at Bolling in a demonstration by the First Experimental Guided Missiles Group.

More than 100,000 spectators were at Bolling that day. They watched two B-29s land after a 24-hour flight from Germany. They saw the dogfighting heroes of World War II — Mustangs — overtaken by a flight of jet-powered F-80 Shooting Stars. (They also saw an unofficial display: Some members of the Air Line Pilots Association took to their planes to write “STRIKE” in smoke over the field.)

Aboard the drone B-17 was Spencer and his crew, along with three journalists who had been invited on the 4 ½-hour trip from Florida. Wrote one of those newsmen in The Washington Post: “From the zigzag takeoff to Washington the drone had smoothly covered the miles without a hand touching controls.”

A little before 1 p.m., the pair of bombers entered Bolling’s airspace and control of the drone was transferred to Hoover, the pilot on the ground. When the B-17 came in to land, it bounced hard on the runway. Like a driving instructor in a car with two steering wheels, Spencer took control, goosed the throttle and lifted the plane back in the air. Ground control again took over and made a second attempt. Same result: a bone-jarring touch-and-go.

From the tower came the command: “Request you make a manual landing.”

Spencer took over. As the Flying Fortress touched down, the right landing gear collapsed and the plane “ground looped” — skittering for more than 1,000 feet, sweeping to the right, its wing digging into the ground. The bomber came to rest on the grass beside the runway. No one on the plane or on the ground was injured.

The Air Force board investigating the incident confronted an interesting question: If it found that pilot error was to blame, who was the pilot? In the 33-page accident report, none of the pilots were blamed. Remote-control landings were inherently tricky, the board wrote. The gear had been weakened by the two earlier hard landings.

Wrote one officer: “On future demonstrations at strange fields the terrain, runway width and length, and wind factors are to be given utmost repeat utmost consideration before a ‘hands-off’ demonstration is attempted.”

Surely everyone involved breathed a sigh of relief. It could have been a lot worse.

Twitter: @johnkelly

For previous columns, visit washingtonpost.com/john-kelly.