In planning its system, Valley Metro is spending extra time, money and attention on safety, saying a lot is at stake: people's lives, the perception of light rail, and money in damage settlements.

Metro engineers traveled to several cities to borrow safety ideas. What emerged were streets designed to separate cars from trains; traffic and pedestrian signals timed to minimize collisions; uniquely engineered rail cars; and plans for a public safety education blitz.

The Federal Transit Administration tapped the state Department of Transportation to oversee safety plans. That double-check was absent in Houston because of a legal loophole, since closed, for systems paid entirely from local funds. Houston rushed to open service in time for the 2004 Super Bowl.

"It's fundamentally programmed for failure because you have people and cars and light-rail trains competing on the same level in the street," said Becky Fenger of Phoenix, who opposed past light-rail propositions. "It will be a bumper-car derby out there."

Tom Bazan of Houston, who has tracked light-rail crashes, said accident data cannot be trusted because it is reported by transit agencies. He accuses operators of reporting low numbers and concealing crash facts, citing missing footage from on-board cameras.

Transit officials say Bazan and his supporters inflate the significance of minor accidents. The FTA tracks only crashes involving injuries or more than $7,500 in damage.

A log of Houston crashes shows three-fourths of them involved automobile drivers making left turns in front of trains or running red lights from side streets. In almost every case, accident investigators blamed drivers. Pedestrian collisions accounted for most of the remaining accidents. FTA officials say the pattern is repeated around the country.

A recent accident in Houston involved a blind man. The man wandered onto downtown tracks in January, then panicked and froze, eyewitnesses said. The braking train struck and injured him.

Metro is using the same predictive traffic-control system as Houston to give trains a jump on green lights and keep them on a reliable schedule.

The Valley's system runs 20 miles in the street and crosses 156 intersections, nearly triple Houston's seven miles and 63 intersections, each an opportunity for a crash.

In 2000, the Insurance Institute for Highway Safety depicted Phoenix as the red-light running capital of the country, with nearly 11 related deaths per 100,000 people from 1992 to 1998. The institute has not updated the study.

The most definitive study of light-rail safety was a 1996 Transportation Research Board survey of 10 light-rail systems. Researchers found that only 30 percent of the track is in streets but that 90 percent of the crashes occurred where cars and trains commingled.

However, the number of crashes is "small when compared to those from typical problem intersections, which usually report about 10 times as many traffic accidents per year," the board's study said.

FTA safety experts say the fatality rate on highways is much higher than on light rail when measured in miles traveled. But in urban miles traveled, light rail killed at 40 times the rate of cars and 10 times the rate of buses, according to an analysis of federal data from 2000 by the Laissez Faire Institute, a Chandler think tank critical of light rail.

Jim Dickey, ADOT's transit division chief, said Metro's work will be scrutinized months before service begins, as empty trains will be tested in the streets. The agency will work with industry consultants and other FTA teams to oversee and revise the safety plan.

This is cache, read story here