(The Globe and Mail posted the following article by Guy Dixon on its website on May 19.)
TORONTO — The sound of freight trains crossing the Prairies or the Canadian Shield evokes another era in transportation but belies the radical transformation that has gone on in railways in recent years.
It may seem like pure common sense, but railways have revolutionized their operations by moving to fixed timetables rather than simply waiting to send trains based on when they have enough cars to make the trip most economical.
“It’s not a new concept in the sense of running to a schedule. What is new is trying to run each individual shipment on that train to a schedule,” said Neal Foot, senior vice-president of operations at Canadian Pacific Railway Ltd.
Today, freight rail is ruled by computer algorithms and centralized data offices, as it tries to compete with trucking to haul specialty cargo such as retail merchandise requiring precise on-time deliveries.
Bulk transports with 100-plus cars all delivering coal, wheat or some other raw good are still an important business for train companies, but it isn’t a growth area.
“The growth rate of bulk, the number of coal mines that are opening up, the number of potash mines opening up are very, very small,” Mr. Foot said.
In 2002, bulk represented 43 per cent of CPR’s revenue, and intermodal container shipping was a growing 25 per cent. The remaining 32 per cent was from carload — a mixed bag of boxcars, automobile transport cars (some specially adapted for sport utility vehicles), centre-beam cars for lumber and other specialized cargo cars.
“It’s the cargo business where we see the growth,” Mr. Foot said, noting that the company’s cargo business grew 9 per cent last year.
“We need the small person who traditionally would have shipped by truck who is now going to ship through a containerized means or on a carload basis by rail. And that’s where we believe our growth is going to be in the future. Some of that will be competition with trucks and some of that will be in partnership with trucks.”
The rationale in sending trains based on tonnage was to avoid trains that were too small to cover costs or to make enough money.
But in the current age of bar codes and must-have delivery times, the drawbacks of tonnage-based freight became too numerous. Shipments were late. Boxcars sat in train yards as they waited to be hooked up to a train. Handling and equipment costs stayed high.
By switching to fixed timetables, Calgary-based CPR found cost savings of more than $500-million since the fall of 2000, around the time the company was implementing its newly scheduled network. Over that period, labour productivity and locomotive productivity rose 40 and 35 per cent, respectively, according to CPR.
“There’s a huge demand placed on us by shippers to be much more responsive, to be much more competitive with the trucking industry, who can provide much more of an individual-load type of focus for them,” Mr. Foot said.
This emphasis on tracking orders from door to door by computers is a far cry from when Oran Hayes used to write down train orders in Capreol, Ont., a junction town north of Sudbury, where he would tell operators where to be and when to wait back in the mid-1960s.
“Instead of issuing train orders by hand, if you will, now the train dispatcher [in Toronto] has to push buttons along the line, and he sees where the train is,” said Mr. Hayes, who is now retired and vice-president of the Northern Ontario Railroad Museum and Heritage Centre in Capreol.
Algorithmic blocks are now used to find the best way to route a shipment. CPR and Montreal-based Canadian National Railway Co., for instance, have used a computer program called MultiRail to help to build new timetables, taking into account so-called “block bypass opportunities” to reduce rail car handling, which slows delivery time.
The program also looks at “circuity,” or the shortest rail distance, and “excessive handling,” which looks for ways to combine other blocks of freight cars to shorten delivery time that much further, while also measuring any congestion of cars in freight yards.
Then there’s the distribution of locomotives along the network, rail line and maintenance planning, crew planning and, finally, a healthy dose of real-world factors from weather to customer behaviour such as incomplete shipping information — all affecting rail times. These factors are then run through computer simulations to help create timetables.
But the algorithms can’t completely reorganize rail networks on their own. “It doesn’t really tell you what to do. It says how good what you said you’re going to do is,” said Mack Barker, general manager of service design at CN.
In addition to its shipment tracking systems and new timetables, he noted that CN has been sharply reducing its number of locomotives and boxcars, further bucking conventional wisdom. Once, more locomotives and more cars were seen as providing more business. Now they can become underused assets, with high labour and maintenance costs.
Since September, 1998, when it instituted its new scheduled network, CN has eliminated 800 locomotives. Its collection of boxcars has declined from 83,000 in late 1998 to 61,500 in 2002, and that number is being further reduced, Mr. Barker said.
Although CPR recently won an award for its operations using the MultiRail software, CN also began using MultiRail in the late 1990s. CN argues that schedule-based management was started around that time by Hunter Harrison, then CN’s chief operating officer and now president and chief executive officer. Mr. Harrison applied the same technique at Illinois Central before that railway was bought by CN in 1998.
MultiRail, which was created by New Jersey company MultiModal Applied Systems Inc. and which costs $2-million (U.S.) to $5-million to implement, helps to simulate which routes require less handling time and, therefore, lower handling costs.