The ATSB says that an air traffic control officer who inadvertently sent a biz jet, a Crown Casino Melbourne Gulfstream IV on a head-on descent in front of a Virgin Australia 737 last year had a ‘mental model problem.’

This thorough and important  report is the most recent of a string of inquiries detailing highly disturbing instances of poorly trained or confused or overworked air traffic controllers allowing passenger jets to suffer breakdowns of separation or ‘near misses’ as they are commonly labelled.

It is a situation that the Minister for Transport, Anthony Albanese, has already acted upon in ordering a special review of such incidents earlier this year, and adds to the urgency with which the air navigation services provider needs to be reformed and properly resourced.

On 8 October 2011 a Virgin Australia Boeing 737-800, which would have been configured with about 180 seats was flying south from Brisbane to Sydney when at a point NE of Armidale it came close enough to meeting a bizjet, the Crown Casino Gulfstream IV, that was descending toward the Gold Coast Airport after a flight from Melbourne that it set off  TCAS alerts in both jets and a system alarm in the air traffic control system.

The diagram above, taken from the report, shows the two stage mental model the ATSB says the controller had for keeping the two jets apart.

However the following graphic, below, shows what actually happened.

The controller had thought he had cleared the G-IV to descend to 39,000 feet until the Virgin jet at 38,000 feet, had passed under it.

In fact he had cleared it to drop down to 31,000 feet, straight through the assigned altitude of the 737.

The report says that when the G-IV pilot correctly read back his instructions to descend to 31,000 feet ‘expectation bias’ on the part of the controller made him think the biz jet had said 39,000 feet as he had intended.

The ATSB report says that AirServices Australia had taken steps to clarify the language and procedures used by its controllers to prevent such  ‘expectation bias’.

The safety investigator clearly doesn’t think the public needs to know more about a peculiar ‘non-operation’ work place event involving the ATC officer concerned 19 days before he screwed up.

This is the reference the report makes:

Non-operation workplace event  

The controller was involved in a non-operational workplace event 19 days prior to the occurrence. The controller attended work the following day, in a non-operational capacity, to ensure that the event was responded to and corrected by management. The controller then took a day’s sick leave before returning to work. The workplace event was reported by the controller to be a period of stress and anxiety.

The ATSB hasn’t included the above for decoration, but relevance, yet the public isn’t told anything more about the non-operational event, which might would allow it to better understand whether AirServices Australia is doing something non-operational which has serious implications for air safety, and whether or not it might perhaps STOP DOING WHATEVER IT WAS.

Did it involve goats, fancy dress, people in drag, or just a group hug-in? Mysterious references likes this do not help the public understand what is otherwise a very useful report, and this is, after all, about the public.