Canadian Transport Lawyers Association Annual Convention
Transportation Safety Board of Canada
at the Canadian Transport Lawyers Association Annual Convention
Data Recorders in Transportation Accident Investigation
September 26, 2008
Thank you very much for the kind introduction. It is a pleasure to participate on this panel with leading transportation specialists. And it is a pleasure to be in Quebec during its 400th year.
Today, I would like to speak to you about data and voice recorders and the essential role they can and must play in accident investigation.
I will look at the patchwork of legal requirements, at some accidents where the investigation suffered from a lack of objective recordings, and at what the TSB has recommended over the years.
Specifically, I will talk about the investigation into the crash of Propair at Mirabel Airport and then about Swissair Flight 111. I will also speak about the Hinton train collision and the TSB's investigation into the derailment of two CN freight trains near Mont Saint-Hilaire. I will end with lessons learned from the investigation into the sinking of the Queen of the North.
Before I get there, I want to tell you a little bit about who we are and what we do. The Transportation Safety Board of Canada, or TSB, is an independent investigative body with one and only one priority - advancing transportation safety.
Our job is to investigate accidents - whether they occur on the water or in the air, or involve Canada's railways or pipelines - and to suggest solutions to improve safety.
We inform the public about what happened, and why. We identify safety deficiencies and make recommendations to government and industry so they may address them.
The TSB is not a regulator or a court; we don't make laws, nor do we find guilt or assign blame.
Approximately 4,000 transportation occurrences under federal jurisdiction are reported to the TSB each year.
When notified of an occurrence, we collect and assess available information, and then decide if a full investigation is warranted. This decision hinges on whether there is significant potential to advance transportation safety.
Where we find risks requiring immediate attention, there are a number of ways in which we communicate those risks. Recommendations are typically used to handle the more difficult, systemic issues.
Now let me talk about recorders. Recorders are an essential tool to investigating transportation accidents. They often hold the clues as to what happened and why in the time before an accident. At the TSB, if we can tell the Canadian public what happened and why, it makes it so much easier to mount a compelling argument for change.
I am not going to address the privilege accorded to voice recordings as Kim is going to cover that in detail. I will focus on the requirements in the modes of transport we investigate and the crucial information we get from these recordings.
With the increasing use of electronic navigation equipment, TSB investigators can also gather information from the memories of GPS units, Electronic Chart Displays and other equipment whenever they are found intact. Interestingly, in a recent rail investigation, we were able to download data from a truck that was stuck on the crossing.
What is clear is that recorded information makes investigations more effective. And we are capable of recording more and more. In my view, the requirements to provide recorded data and voice to investigators should keep pace with technological advances.
They should also be the same in all modes and the domestic requirements should be as stringent as the international requirements.
To underscore this point, let's look at a number of high profile investigations in the last 25 years. I will talk about the information available to investigators, and the information that was not available and the difficulties this presented to each of these investigations. And I want to end by telling you about the way forward and what the TSB has recommended in the area of recorders.
The aviation sector is the leader in data recorders. For over 40 years, air investigators have relied on recorded voice and data from the so called "black boxes." From the picture you will notice that black boxes are not actually black nor are they always box shaped.
The Canadian Aviation Regulations require cockpit voice recorders and flight data recorders on many aircraft. They also contain specific information as to recording times and survivability requirements.
CVRs record radio communications of the flight crew, interphone conversations between the cockpit and cabin, and ambient sounds within the cockpit.
FDRs record a number of aircraft parameters depending on the recorder, such as airspeed, altitude and heading, flight control and engine thrust lever positions, navigational information and aircraft systems information.
Let's take a look now at an accident where the on-board recorder made the investigation more effective. On June 18, 1998 a Metro II crashed at Mirabel Airport killing all 11 people on board.
The investigators learned from the physical wreckage and the air traffic control tapes some of what happened to this aircraft.
However, the voice recorder allowed them to piece together the chronology of events - providing a clear picture of the crew's actions during the developing emergency.
This clear picture allowed the Board to make recommendations for change.changes that the regulators implemented.
Now on to an example of an investigation in which we had recorded information but not everything we needed.
The 2nd of September of this year marked the 10th anniversary of the night Swissair 111 crashed off Peggy's Cove.
On September 2, 1998, Swissair Flight 111, a McDonnell-Douglas MD-11 departed New York City on a scheduled flight to Geneva, Switzerland, with 215 passengers and 14 crew on board.
About 53 minutes into the flight, the crew smelled an abnormal odor in the cockpit. Their attention was drawn to the area behind and above them and they began to investigate the source. After further troubleshooting, they assessed there was definitely smoke and decided to divert to Halifax.
While the flight crew was preparing to land, they were unaware that a fire was spreading above the cockpit ceiling. Soon thereafter, the aircraft's FDR logged a rapid succession of system failures. The crew declared an emergency and an immediate need to land.
About one minute later, radio communications and radar contact were lost, and the flight recorders stopped functioning.
About five and a half minutes after that, the aircraft crashed into the ocean with the loss of all 229 souls on board.
This tragic accident led to a four and a half year investigation - the largest and most complex ever undertaken by the TSB.
The TSB made 23 recommendations as part of the Swissair investigation. In eight of those recommendations, or about a third, we called for improvements to recorders aboard aircraft. Both the cockpit voice recorder and flight data recorder on Swissair 111 ceased to function some five and a half minutes before impact.
Consequently in this investigation, as with many others, the lack of quality data severely hampered our efforts to validate some of the primary safety deficiencies.
The TSB's recommendations on recorders included:
- increasing recording capacity;
- increasing the quality and readability of data;
- making sure they continue recording in the event of electrical failures;
- requiring image-recorders; and
- protecting the confidentiality of cockpit voice and image recordings.
So what has been done to address data recorder deficiencies?
In March 2008, the FAA issued new regulations stating that by April 2012, CVRs on all turbine engine-powered airplanes:
- must have a 2 hour recording capacity;
- must have an independent power supply that provides 10 minutes of electrical power; and
- any single electrical failure must not result in disabling both the CVR and the FDR.
This is real progress, and addresses three of our recommendations.
But there are some notable outstanding issues regarding recorders after Swissair - 10 years later.
If you were to go back to the investigation report, you would note that our recommendation on image recorders is explicitly paired with our recommendation on harmonizing the international treatment of cockpit voice and image recorders. We continue to believe that image recorders are essential to modern accident investigation.
In the rail mode, recorders have been a long time coming.
On February 8, 1986, near Hinton, Alberta, a freight train collided with a passenger train. 23 people died and a Commission of Inquiry was called to find out why. The best available evidence came from the surviving crew member.
However, when the two trains collided he was in the caboose - more than a mile away from the head end. In addition to the issue of physical proximity, there were many inconsistencies with his evidence and the Commission had no choice but to give it little weight.
The two men who knew most about what happened - the locomotive engineer and the trainman, both died. Aside from the overwhelming evidence that they had proceeded through a stop signal, there was no other direct evidence of the circumstances in which this took place or "why" it happened.
In the absence of this evidence, the Commission tried to learn everything they could about rail operations and the work patterns and histories of the crew to build a theory from the available evidence. This painstaking work took close to a year. It included 48 days of Public Hearings in which 150 witnesses were heard. In the end, Mr. Justice René Foisy had a picture of what might have happened. But we will never know for sure as the freight train was equipped with neither a data nor a voice recorder.
With each investigation we learn more about what we need to do a thorough job. Today, the Railway Locomotive Inspection and Safety Rules require locomotive event recorders.
All locomotives, except those operating in rail yards, are required to have one. Locomotive event recorders log information such as date, time, speed, throttle position, brake operation and horn and bell use.
This is progress compared with the Hinton days but we still have no voice recorders on locomotives as we have on aircraft and the recorders we do have will not always survive the horrific forces of an accident.
One such accident happened on December 30, 1999. A CN train loaded with hydrocarbons was travelling westward on the north track of the St-Hyacinthe subdivision. Near Mont St-Hilaire, Quebec, some cars derailed and blocked the adjacent south track. Another train travelling eastward on the south track collided with the cars of the westbound train as they derailed. A total of 60 cars derailed and most were destroyed.
In this time, the Canadian public continues to be exposed to risk that could be avoided. As I said before, the slow pace of regulatory change can sometimes be a safety issue in and of itself.
There was an explosion and some cars burned for more than four days, creating a smoke plume about 500 metres high. The two crew members on the eastbound train perished in the accident.
Approximately 2.7 million litres of gasoline and heating oil spilled and caught fire, damaging private property and the environment.
The event recorders from the eastbound train were destroyed by fire. That is why the Board recommended that locomotive event recorders be both crashworthy and fire resistant. This would bring them in line with recorders on aircraft.
We are pleased we now have survivability requirements for locomotive event recorders. It is disappointing that these only apply to recorders installed in new locomotives because locomotives currently in service will be out there a long time.
We would like to see crashworthy event recorders installed on all locomotives, and we would like voice recordings - so that the best data will always be available to accident investigators.
Now let's talk about recorders in marine transportation and the sinking of the Queen of the North.
In the early morning hours of March 22nd, 2006, the Queen of the North en route to Port Hardy from Prince Rupert, BC, struck the northeast side of Gil Island. The vessel sustained extensive damage to its hull, lost its propulsion, and drifted for one hour and 17 minutes before it sank in 430 meters of water. Passengers and crew abandoned the vessel before it sank. Tragically, two passengers were lost and have since been declared dead.
The Queen of the North did not trade internationally and she did not have a Voyage Data Recorder, or VDR.
Canada is a signatory to the International Convention for the Safety of Life at Sea. This means all Canadian vessels built after July 2002 and all Canadian passenger vessels that trade internationally must carry a VDR. They log the ship's position, speed, heading, bridge audio and radio communications, and radar data. VDRs also record information about ship systems and controls.
Simplified VDRs are to be installed on cargo vessels built before 2002 that trade internationally. They record the same data as VDRs except the controls and systems information.
Now back to the Queen of the North. I want to talk about the data we were able to retrieve and about the information that was not available and what that meant to our investigation.
We conducted a series of dives to the wreck to retrieve bridge navigational and communications equipment and recover important data.
This allowed us to reconstruct the track followed by the ferry that night.
What you are about to see is a time-compressed simulation of the voyage of the Queen of the North. It is based on the data recovered from the bridge equipment, and mathematical models of the vessel tested on a bridge simulator. You will see screen shots from the radar and the electronic chart system display, showing the vessel's intended and actual tracks. The conversations you hear are between the bridge crew and marine communications and traffic services. (Play Video)
From this, we were able to determine that sound watch keeping practices were not being followed. As such, a critical course change was not made and the vessel continued on its course until it struck Gil Island, suffered damage and sank.
Because we did not have a VDR, including bridge audio, we could not determine the exact circumstances leading to the missed course change. The question in many peoples' minds was: "what happened in those final 14 minutes?" We may never be able to answer this question with finite detail or absolute certainty.
The final investigation report included three recommendations. One of these called for VDRs on ALL of Canada's large vessels.
Recorders are essential to conducting successful and thorough accident investigations. Some good work has been done to improve the quality and availability of data recordings in all the modes, but more needs to be done so that quality data is available to investigators following an accident.
After Swissair, we made recommendations calling for the installation of cockpit image recorders.
Understandably, crews oppose additional surveillance in their workplace, but this additional information would help investigators better understand the events leading up to the accident.
This resistance can only be overcome if the international community protects the confidentiality of all recordings. We must ensure they will not be released and will only be used to advance transportation safety. We have also called for audio recording capabilities aboard locomotives to assist in our rail investigations.
Rail and marine transportation also need to catch up to aviation in their use of modern recorders. While some marine operators have installed VDRs in domestic vessels, we would like to see all large domestic vessels equipped with VDRs.
We would also like to see crashworthy Locomotive Event Recorders installed on all Canadian railway locomotives.
Thank you for your attention.
- Date modified: