/

Programming error suspected behind scrubbed rocket launch

Kyodo

Experts suspect a computer programming error and lax preliminary checks were among the reasons Japan’s newest rocket didn’t get off the ground Tuesday.

The computer controlling the launch of the three-stage Epsilon rocket at the Uchinoura Space Center in Kagoshima Prefecture detected an abnormality in the rocket’s position only 19 seconds before its scheduled liftoff scheduled for 1:45 p.m., but it was later found to be normal.

Experts believe a computer programming error and lax preliminary checks were behind the trouble.

“It may have been an elementary, but not serious, problem,” one expert said.

The followup inspection found no abnormality with the attitude sensors in the rocket or with the computer feeding the data to the ground, according to the Japan Aerospace Exploration Agency (JAXA).

The agency said the computer on the ground is tasked with determining whether data sent from the rocket are within acceptable parameters for a launch.

JAXA officials said at a news conference Tuesday that the data flow between the computers on the rocket and on the ground went wrong but declined to give details.

“There may have been a problem with the program of the computer on the ground for determining the rocket’s attitude,” speculated Harunori Nagata, a professor of space propulsion engineering at Hokkaido University.

Tetsuo Yasaka, a professor emeritus of space engineering at Kyushu University, did not mention specific causes but said that “it is common that signals from the rocket are not reflected on the ground system due to some discrepancies (between computers).”

Aborting a launch is not uncommon for newly developed rockets. Japan’s H-I and H-II rockets also experienced aborted launches.

Yasaka said this type of problem can be quickly detected through a careful inspection prior to the launch.

“The fact that the (problem) came out at this stage (just before the launch) is not really encouraging,” he said.

“The (full) investigations of the cause are yet to come,” JAXA official Akira Konno said in Tokyo. “I believe it is likely that the computer on the ground had a problem. The timing for receiving the rocket’s data on the ground may not have been correct.”

The Epsilon project involves the use of artificial intelligence for prelaunch inspections. Yasuhiro Morita, a JAXA professor who was in charge of developing the AI, said “the latest problem has nothing to do with the AI.”

Asked whether there will be a long delay until the next launch attempt, Yasaka said that “it should only require a minor repair because possible causes have been narrowed down.”

Nagata sounded an optimistic note, saying that “none of the equipment, including the launchpad, the satellite and the rocket, has been damaged. It (the postponement) does not mean a failure.”

Hiroki Matsuo, who formerly headed an education and science ministry committee on space development, said that if JAXA’s explanation is correct, the repair may be relatively easy. But he added that those involved in the project “must really think about why they overlooked (the problem) until the day of the launch.”