The UK's Rail Safety & Standards Board - RSSB - published its annual safety performance report for 2009/2010. I read the "Key facts and figures" bit - and fascinating it was too.

There were no passenger or workforce fatalities in train accidents in Great Britain during the year.

There were five passenger fatalities (two on platform edges hit by trains, two falling off platforms and one falling on an escalator): this is equal to the lowest annual figure on record. There were three workforce fatalities, and 298 public fatalities - suicides, trespassers or people misusing level crossings.

Most of the risk of fatalities on the railway is through the behaviour of members of the public.

The annual average number of workforce fatalities has dropped by 98% since 1949: passenger fatalities are also significantly down. Public fatalities have shown no significant change.

In the last three years, and in five of the last 10 years, there have been no fatalities to passengers in train accidents. The moving average number of train accidents resulting in passenger or workforce fatalities has fallen by 75% since 1950, and is currently less than one a year.

These are excellent figures, superb figures, and it is just a shame that, because they are good news, they hit no headlines. Because the rail system in Great Britain carries around 24,000 trains a day with trailing loads of up to 1000 tonnes: they travel at speeds of up to 200 km/h and are powered by either high-voltage electricity or flammable diesel fuel.

They are driven, controlled and signalled by fallible human beings. Several stations see more passengers than Heathrow Airport.

And yet no passenger has been killed in a train accident for three years: there were no such fatalities in half of the last 10 years.

Not only is that good news, it is a high tribute to the excellence of our rail system - all of it. Managers, workforce, maintainers, everyone involved.

The report can be found on RSSB's web-site, www.rssb.co.uk.

An IATA official has recently been quoted as saying that, 'A slot management system is indicative of a failure of governments or airports to invest in adequate infrastructure to keep pace with airline demand.'

Should governments and airports always invest to meet airline demand?

Consider the issue of de-peaking. Until recently, airlines worked on a hub-and-spoke basis, scheduling large numbers of flights in and out of hub airports in 3-4 waves each day - early morning, lunch time, late afternoon, and evening. They all connected with each other and then flew off again.

Surprise, surprise - this caused overloading of facilities at peak times and under-utilisation out of those peaks. What did airlines do? Why, they de-peaked at key airports.

Aircraft still came into hub airports in those waves, but their departure was sometimes re-scheduled so that the peak load on check-in and baggage staff was reduced.

So - by IATA's argument - airports should have invested in capacity for the peak caused by the historic hub-and-spoke operation when that became fashionable. They would, of course, then have been left high and dry by the subsequent de-peaking, having invested in long-life assets which the users no longer wanted!

Consider the fact that aircraft are mobile assets and airports are not. Some airlines will - rightly - try a route for a season or two to see if they can make money out of it. If they can't, they pull out. So what should the airport do? How much should they invest for an operation which might not last beyond tomorrow?

And consider the impact of frequent flights by small aircraft or less frequent ones by larger aircraft (an issue studied, incidentally, in ACRP Report 31, available on the TRB web-site).

If an airline schedules a 50-seat plane every half hour instead of a 100-seat plane every hour, it takes up twice as many slots for the same number of passengers. OK, on this one you could argue passenger convenience - all other things being equal, people prefer frequency. But is it really worth it - is it a good use of scarce runway and slot capacity?

Is it worth increasing capacity - as IATA suggests - for this kind of thing?

To me, the killer argument here is that of San Diego. This is the busiest single runway airport in the United States, handling 17 million passengers a year. It has spent much time, money and effort looking for ways to increase capacity because, it is said, it's full.

But look across the Atlantic, to Gatwick - the busiest single runway airport in the world, handling exactly twice as many passengers! The key difference is aircraft size.

The predict and provide policy - predicting demand then providing infrastructure to meet it - has been discarded in other areas. How much does IATA need to catch up with this? Because another leading IATA manager has been quoted as saying that airports should use their charging system to re-distribute traffic - a peak charge signals the cost of the additional infrastructure needed for the peak.

A colleague recently moved from the aviation industry to the rail industry. The key difference he noticed was that the former was always planning for growth, whereas the rail industry was planning for a steady state.

This led me to wonder what the aviation industry would look like if it wasn't always growing! I have as yet come to no conclusions, but there was an interesting presentation at the TRB Annual Meeting in January which showed that air passenger growth might be slowing significantly.

And, logically, it cannot continue. In my lifetime, fares have continued to drop in real terms, and this has led to the growth in air travel we have seen. Can this go on?

Your thoughts would be appreciated!

I was very interested to read 'The New York City pedestrian safety study & action plan', published by the Department of Transport this month - see the New York City web-site.

Headline facts are that
pedestrians are 10 times more likely to die than a motor vehicle occupant in the event of a crash,
pedestrians accounted for 50% of traffic fatalities in 2005-2009,
36% of crashes resulted from driver inattention,
27% of crashes where pedestrians were killed or seriously injured occured when a motorist failed to give way,
80% of crashes which kill or seriously injure pedestrians involve male drivers, and
79% involve private cars.

New York City is relatively safe as American cities go - there are 3.49 traffic fatalities for each 100,000 of population (compared with Atlanta at 10.97 and Detroit at 10.31). But London has 2.10 and Stockholm 1.23. Since national fatality rates are 12.2, it must be pretty deadly out in the sticks!

Since 2004, the number of traffic fatalities in New York City has been lower than in 1910 - which is impressive!

Pedestrian fatality rates were 1.8/100,000 of population in 2008 - again much lower than Atlanta (11) and Los Angeles (7.6).

Vulnerable road users - pedestrians, cyclists and motor-cyclists - accounted for 71% of the city's traffic fatalities.

A telling quote from the report is that, 'Drivers can usually be expected to behave in ways appropriate for their own safety but may not always have the same commitment to the safety of people outside vehicles'.

A disproportionate number of pedestrian crashes in Manhattan were on major 2-way streets.

Wide roads also accounted for more pedestrian crashes - this reflects much other research which shows that widening roads reduces safety.

74% of crashes involving pedestrians were at intersections (it's 24% in the rest of the USA). Nearly half were at signalled intersections: 57% of these occurred when the pedestrain was crossing with the signal.

Pedestrians are more likely to be killed or seriously injured when crossing with the signal than against it, although crossing against the signal is more likely to be fatal.

Driver inattention was a factor in 36% of crashes: these were more than twice as likely to be fatal than others. Driving while using a mobile phone is more dangerous than drunken driving.

Speed is also an issue. A pedestrian struck at 40 miles/h is 4 times more likely to die than one struck at 30: one struck at 30 is 6 times more likely to die than one struck at 20.

As a result of the study, an action plan has been drawn up. High-crash corridors are to be identified and targeted, and countdown signals are to be installed at more crossings. The State legislature is to be pressed to allow more red light cameras (only 150 are allowed at the
moment) and speed cameras (currently illegal in New York City).

The city's target is to halve traffic fatalities between 2007 and 2030: they are currently a bit ahead of traget, with 256 fatailites in 2009.

Airports Council International produce a series of best practice guides based on their regular Airport Service Quality benchmarking exercise. The one on Baggage Carts (trolleys), published in 2008, has some interesting statistics.

Airports where carts are free to use score higher on "Availability of baggage carts" than those which charge or ask for a deposit.

90% of airports have carts which can be used in lifts, 62% on moving walkways, and 12% on escalators.

44% have brakes, although the report doesn't say what sort.

The usual ones at airports have brakes which are only on - applied - when a passenger activates them, usually by holding two parts of the handle together.

By contrast, carts used on railway stations normally have the brakes on unless someone holds the two parts of the handle together: only then is the brake off and the cart can move. This, obviously, is a safety measure - no-one wants baggage carts on a railway line!

But it does pose a problem when the station is at an airport. What kind of cart do you use? If you use one where the brake isn't normally on, then it can't be safely used on the station - passengers can't take them to the train door.

An issue for airport railways!

What happened yesterday.
I got to my local station yesterday morning to find a crowd round the secondary entrance, the one I usually use. There was a barely audible announcement going on. The ticket barrier gates were closed and no staff were visible: there were a few people actually on the station.

Passengers being passengers, they had found that one of the barrier gates - the extra wide one for wheelchairs, buggies and large cases - could be pushed aside enough to allow passage, so a trickle of passengers was getting onto the station.

Then there was an announcement which I could only just hear to the effect that the next southbound train was at Luton, 16 km to the north, and there was an implication that it was being held there. I also caught something about a lineside fire.

I went towards the main entrance and the bus station: as I did, people were coming the other way saying that the main entrance was shut and they were hoping to get in by the secondary one.

I sent a quick text message to my daughter, who also commutes.

As I got to the bus station I saw a bus labelled "Stevenage" loading up: I joined the queue and got a seat. I reasoned that it would go to anything between one and three useful stations. When it had its full load of seated and permitted standing passengers, it set off: several were left behind.

It took half an hour to reach Hatfield station, some 10 km away.

Train passengers dominated: there were very few left on the bus as we got off at Hatfield, and only a few had boarded or alighted along the way.

At Hatfield, I caught a train to London. The service isn't nearly as good as at my local station: there was an 8:02 calling at most stations to Moorgate then an 8:18 semi-fast to King's Cross.

I got off at Finsbury Park and used the Victoria and Bakerloo Lines of the Underground to get to work at Paddington: the journey took 40-50 minutes longer than usual.

The use of barrier gates.
I did ponder the use of barrier gates to bar access to the station in times of disruption.

It wasn't working on the unstaffed side of the station. Determined passengers could by-pass them relatively easily - although presumably where there were staff, people were not getting in.

What are the advantages of trying to keep intending passengers out?

There may be a long-term issue of over-crowded platforms - although since there were few passing trains, the danger was limited!

Fewer staff would be exposed to the wrath of the frustrated Monday morning commuter - a force to be reckoned with!

Some passengers might have got onto the platforms intending to travel by train and then changed their minds: if they had day tickets (as opposed to season tickets) this might have caused problems.

There may be advantages in keeping passengers off the active railway.

Against this is the problem of communication. Station public address systems are designed to be heard by people on stations, not in the ticket barrier area (or near it, because of the crowds). So communication of the problem, of news and of updates was more difficult, which no doubt led to more frustration.

Another problem is that the gates would have had to be opened some time. Passenger psychology - especially commuter psychology - is that they have to get to the right platform quickly. There would be no orderly walk along the platform to the footbridge, over the footbridge and down to the London-bound platforms: there would have been a scrum.

The probability that someone would get hurt in the rush was quite high!

This depends to a degree on the length of the delay, the advice given to passengers and the alternatives available.

The worst combination is probably a short delay and no advice or information - so crowds would build up. As time goes on - as delays lengthen - it is more likely that the train operator would be able to find buses to take people to working stations, and passengers are more likely to find their own way (they can be a resourceful lot!).

It was an interesting experience, and one which will feed into IARO's forthcoming report on disruption, in course of preparation.

Thoughts on this use of ticket barrier gates would be welcomed.