With his body perched precariously on the side of a rubberised boat, and one hand clutching a hydrophone, Mr Bryan Siow strains to listen for signals in the water.
As he handles what is basically a souped-up microphone designed to pick up sounds underwater, the senior air safety investigator also has to deal with the bobbing waves, taking care not to fall overboard, or drop his equipment into the sea.
"It's always challenging when you work off a small boat," he told The Sunday Times. "Even when the waves are not heavy, the motion tends to reduce your effectiveness when you're trying to listen for what is quite a faint sound."
Mr Siow, 38, is one of three investigators from Singapore's Transport Safety Investigation Bureau (TSIB) sent to join Indonesia's national transportation safety committee KNKT in the open-water search for the "black box" recorders from Lion Air Flight JT610.
It is a gruelling mission that could go on for days, but a critical step in the probe for answers on why the relatively new Boeing 737 Max-8 passenger jet, with 189 people on board, plunged into the Java Sea on the morning of Oct 29.
The TSIB team, led by senior investigator and the bureau's head of operations Steven Teo, was mobilised after KNKT accepted Singapore's offer to help in their search.
The team was en route to Jakarta within hours of the crash, arriving at about 7pm local time, before being whisked away to rendezvous on the research ship Baruna Jaya 1 with their KNKT counterparts.
At midnight, they set sail together for the crash site in waters off Karawang regency, in West Java, fully aware that they might be out at sea for some time to search for the flight data recorder (FDR) and cockpit voice recorder (CVR) from the ill-fated flight.
The search began at sunrise but it did not take long for the teams to catch a lucky break. They picked up the first 'ping' - a short high-pitched signal - from an underwater locator beacon attached to one of the black boxes, on the first day of their search with the KNKT.
Recounting how it happened, Mr Teo, 40, said KNKT and TSIB investigators had earlier agreed on a search area with eight points, four on each side of a grid based on the flight profile of Flight JT610 before contact with the plane was lost.
"The aircraft seems to have continued its flight profile," he added. "And once we reached the first point, we could hear the ping."
The next day, investigators from both agencies managed to narrow down the estimated location of the signal source from 200m to 20m from where the FDR was later recovered by divers on Nov 1.
That was no mean feat, considering the radius of the search and rescue operation had stretched to 15 nautical miles (27,780m) by then.
Also, despite having a rough bearing on the signal, Indonesian divers still had to manually comb the area to find the source of the ping, and eventually the FDR.
"And it was not in plain sight, it was actually buried," said Mr Siow.
Indeed, dive teams reported encountering mud of up to 1.5m deep on the seabed, which not only hampered their movement but also affected their visibility underwater.
"So, really, our hats off to the divers who found it," said Mr Teo, describing their job as not unlike "looking for a needle in a haystack".
EVERY CRASH HAS ITS CHALLENGES
The TSIB search team, along with a fourth investigator, Mr Ng Junsheng, met with The Sunday Times last Monday when the Baruna Jaya 1 returned to port for resupply.
Mr Ng, 37, who is also head of technical support at TSIB, had arrived in Jakarta on Nov 1, to join KNKT in processing the flight recorder retrieved by divers earlier that day.
It was the first time Mr Siow, Mr Teo and Mr Alexander Leong, 34, had returned to shore since the night of the crash.
During the operation, the team would set out from the Baruna Jaya 1 in smaller rigid inflatable boats, or RIBs, and deploy their underwater locator beacon detector system, alongside KNKT officers.
"Our longest stretch at sea actively searching for the recorders on a RIB was about four hours," said Mr Teo, who has been an air safety investigator for almost eight years.
For the search, TSIB had sent two sets of the underwater locator beacon detector system, which comprises a hydrophone that "listens" for the ping, and allows its user to triangulate its approximate location with a ruggedised tablet computer.
"The system depends on the human ear, so sometimes when we get a signal, we may also want a second confirmation, so between ourselves and KNKT, we take turns to listen for the ping and work as a team to determine the direction where it is coming from," said Mr Teo.
The day after their interview, Mr Teo and Mr Leong were out at sea again to search for the CVR.
Mr Siow, an air safety investigator for about 10 years, was involved in a similar operation during the 2014 crash of Air Asia Flight QZ8501 in the Java Sea. He says there is no telling when the CVR will be found this time round.
"Every crash comes with its own challenges and if you compare this crash with QZ8501, the recorders then were found a day apart from each other," he added. "But for this incident, we are still looking for the second one."
With the exception of Mr Leong, who joined TSIB three years ago, the rest of the team were also involved in the operation to recover the Air Asia flight recorders. All agreed that weather conditions were more favourable this time.
"For this mission, they can reach the search location in three hours, but for the QZ crash, they needed to travel for at least 14 hours," said Mr Ng.
"QZ also happened in the middle of the monsoon season and so for safety reasons, it took the team quite a few days before they could deploy out to sea."
EVOLUTION IN AIR CRASH PROBES
When asked if he ever thought he would spend so much time at sea even though he is an air safety investigator, Mr Siow said he did not.
"But after the crash of Air France 447, the investigator's role seems to have evolved, and we are now very focused on sea operations," he said, referring to the Paris-bound flight that crashed in the Atlantic Ocean in 2009, killing all 228 on board.
"Even the training has evolved to focus on the search for flight recorders at sea, not just on land."
All four TSIB investigators are back in Singapore as of yesterday, even as the Indonesian authorities called off the search for crash victims after almost two weeks.
Mr Teo said TSIB has a close working relationship with KNKT, and besides search and recovery operations, they also cooperate in training, often sharing investigation techniques and lessons learned.
The TSIB also works with its other international peers, and is often sought out to assist in the search for black boxes at sea and to recover data from flight recorders.
For instance,Mr Teo was involved in a search operation in South Korea, after an Asiana Airlines cargo plane went down in waters off Jeju island, in July 2011.
"It's always easy to make mistakes, and it's a lot harder to do it right," said Mr Ng, when asked how it was like when tasked with processing data read-outs from recovered flight recorders, which are crucial in any air crash investigation.
"They have placed a fair amount of faith in you," he said. "So you really want to ensure you did a thorough analysis, that the data is accurate, and you were able to support the wider investigation in a meaningful way."