On the 30th anniversary of America's worst commercial nuclear accident, one thing is clear about the impact Three Mile Island had on nuclear power.

TMI's legacy is an epic tale of technological innovations overcoming technological mistakes.

The partial meltdown of the nuclear core in TMI's No. 2 reactor on March 28, 1979, near Middletown, Dauphin County, caused no deaths and resulted in a relatively slight, albeit notable, belch of radiation.

But for five days, residents of Central Pennsylvania and throughout the world received a media barrage of alarming stories about efforts to prevent radiation from spewing from the power plant.

International interest and growing concern even prompted President Jimmy Carter, against the advice of his staff, to arrive at Middletown to help then-Gov. Dick Thornburgh gather information about the extent of the crisis and the degree of danger.

The most serious accident in U.S. commercial nuclear history, Three Mile Island halted construction of nuclear plants, heightened regulatory oversight and spurred sweeping changes in nuclear power plant operations.

Today, debate continues about how the accident shook public confidence to a point where it still affects the industry. The fear it generated stunted the nuclear power industry growth for decades.

"The response by the media, state agencies and everyone involved scared the public, and that trauma is still evident today," said Michael D. Pavelek II, a native of Allegheny Township, Westmoreland County, who was a superintendent in the TMI-2 cleanup. "That trauma is what set this country back in nuclear power by 30 years."

Elizabeth Hayden, spokeswoman for the Nuclear Regulatory Commission, agreed that "TMI got people thinking."

"We have enough [electricity production] for current demand, so everything came to a screeching halt," she said. "Then the NRC tightened the regulations so we wouldn't have another TMI."

But there's also this argument: TMI forced an overly confident industry to re-evaluate its safety record, which prevented worse accidents.

Among the improvements: TMI brought better computer monitoring of nuclear power plants, better simulators to train personnel and simpler, safer plant designs. It also prompted formation of the Institute of Nuclear Power Operators and the international version, World Association of Nuclear Operators, to which every nuclear utility belongs. The organizations monitor plants and help improve operations.

As a result, the 104 nuclear power plants in the United States and the 440 worldwide have helped restore public confidence and put the industry on the cusp of renaissance.

Larry Foulke, former president of the American Nuclear Society and now director of nuclear programs at the University of Pittsburgh's Swanson School of Engineering, said the accident led to lessons learned and improvements in power plant technology.

"But it also led to a changing of attitude in the industry," he said. "Before TMI, the industry was sort of smug with attitudes that it couldn't have an accident. After TMI, that attitude changed from smugness to vigilance."

Robert L. Long, vice president of TMI's operator, GPU Nuclear, and now owner of Nuclear Stewardship LLC, a consulting company, said he and others who'd spent careers in nuclear power "didn't think we could have a core meltdown."

"We thought the system could handle the worst accident -- a loss-of-coolant accident," he said. "It was hard for all of us to think that severe damage had occurred in the core until we looked at it when it became a reality."

The 1986 Chernobyl reactor explosion in northern Ukraine, considered to be the worst nuclear plant disaster, bolstered global fears by proving that nuclear accidents could cause death and natural destruction.

Still, TMI did show that containment buildings worked, meltdowns could be contained and radiation levels could be controlled. Also, the TMI-2 core actually never melted, Mr. Pavelek said. Instead, it collapsed into rubble.

NRC statistics show that the number of significant events has dropped from an industry average of 0.9 per year in 1989 to 0.01 per year in 2006. A significant event affects the performance of a plant, increases the probability of damage to its core or causes an abnormal occurrence in plant operations.

In the end, the TMI accident, meltdown or not, did produce a major meltdown of public confidence in nuclear technology.

"We're suffering from the impressions of TMI rather than the realities of TMI," Mr. Pavelek said. "The perceptions are what we have suffered from.

"But we've had a long, long time with it, and we've learned a lot," he said.

The TMI accident occurred when a valve in a pressurizer, which regulates pressure inside the reactor system, failed to close, giving technicians inadequate information and leaving them unable to understand what was happening.

They thought the pressurizer was becoming dangerously full of water, when it actually was leaking. Mistakenly, they reduced water flow into the pressurizer, thus reducing available water to cool the reactor, which began to overheat.

"We worked long, long hours, but we knew what the radiation levels were and knew there was no [immediate] danger," Dr. Long said. "We were trying to help the operators and work with all the consultants to figure out how to get the plant under control."

But in the end, control room operators made errors that melted half of the uranium fuel pellets.

"The operators almost had too much information," Dr. Foulke said. "Hundreds of alarms were going off and hundreds of lights were flashing" without providing "any useful information."

The initial mistake prompted a domino effect of errors before the plant's operators shut it down. It led to redesign of control rooms equipped with computers that prioritize problems and eliminate the "avalanche of alarms." Nowadays operators monitor readings but allow computers to fine-tune operations.

Soon after the TMI accident began, people throughout Central Pennsylvania and as far away as China began fearing imminent danger from radiation exposure.

Dr. Long and other utility officials came from New Jersey to the island in the Susquehanna River, where TMI is situated just south of Harrisburg, to help understand the problem.

"What impressed me the most, these men were struggling with a very difficult situation and working calmly to understand what was going on," he said. "Nothing seemed to work. Procedures didn't apply to what they were seeing and the behavior of the plant.

"They never gave up and kept going and were not frightened. They were well trained. Unfortunately their training led them to make mistakes that caused the accident."

As history has proven, the NRC also was wrong in early public announcements that radiation levels from TMI indicated a ruptured reactor vessel. Officials obtained radiation levels while flying over the plant two days after the crisis began, but later misinterpreted those readings.

It didn't help that "The China Syndrome," a movie released days before the accident, seemed so prophetic. That timely plot involved a core meltdown in a California power plant, with one character explaining the "China Syndrome" -- a theory that a reactor fuel meltdown could burn through the earth to China -- before speculating that the cinematic accident would render an area the size of Pennsylvania permanently uninhabitable.

Next, the NRC said a hydrogen bubble inside the TMI reactor might explode. Not true, as it turned out. But an already rattled public waited with growing trepidation as experts, working to stabilize the situation, quit trying to correct media doomsday reports that fueled public panic.

Concern grew to a point where then-Gov. Dick Thornburgh recommended evacuation of pregnant women and preschool children from the area within five miles of the plant. About 140,000 people fled for up to a week, when conditions stabilized at the plant and officials no longer believed an emergency was imminent.

Cleanup posed new concerns. GPU Nuclear feared exposing cleanup crews and the public to radiation during the cleanup -- a process that required careful planning.

It took two years before GPU dropped a camera inside the reactor core, Dr. Long said, which proved that nuclear fuel had been destroyed.

Mr. Pavelek, who spent those two years determining the conditions inside the containment building, was involved in proposing the use of robots in the cleanup of the containment building basement, where humans could not survive the deadly radiation dose.

When GPU sought a supplier, Carnegie Mellon University roboticist William "Red" Whittaker saw an opportunity, noting GPU's "concerns of failure, of doing wrong things and looking bad."

Dr. Whittaker stepped in with plans for robots to replace vulnerable humans in remote, potentially deadly environments -- including radioactive containment buildings. He and his students produced wood and plastic robotic models, then drove with them to TMI to convince GPU officials of their potential.

"I saw in a moment what could be done," Dr. Whittaker said. "This presented tremendous prospects for a robotic solution."

With an OK to proceed, his team of two dozen students worked under tight deadlines to produce three robots.

The first, "Reconnaissance," completed within a year, was put into action in 1984, five years after the accident, creating "a major leap" in robotic technology, Dr. Whittaker said. It served cleanup duty for four years inside the containment building basement, then was succeeded by "Core Sampler." A third robot, "Work Horse" was built but never used.

The robots, which were submersible in water, were made of materials that wouldn't absorb radiation and lacked crevices where radioactive dust could accumulate.

Operated by humans at a distant control panel, the robots had to unroll enough electrical cord to provide power hundreds of feet down corridors deep inside the containment building. It was a major engineering feat that Dr. Whittaker initially designed on a napkin.

The robots carried cameras and equipment to measure radiation. They also bore high-pressure water sprayers, core drillers and scrapers.

Over the years, CMU's two robots cleaned up enough radiation in the basement to allow humans briefly to enter and evaluate conditions. A vacuum system was used to remove the damaged nuclear fuel from the reactor.

Still slightly contaminated, the containment building will remain sealed until TMI reactor No. 1, which is still operating, is decommissioned.

The $1 billion cleanup took 14 years to complete. Without the CMU robots, GPU would have been forced to recruit workers for short stints of cleanup duty at high cost and health risk.

"His team made the difference for us, when you look at the contribution it made in terms of human exposure and money to train people and bring people in," Mr. Pavelek said. "When you look at the money and expense [CMU saved], it was phenomenal.

"I have to tell you, I was amazed at the talent," he said. "[Dr. Whittaker] is an exceptionally intelligent man, but the team of students he put together astonished us."

Dr. Whittaker called it "a galvanizing experience that seeded the robotics movement."

"Before this, there was speculation on what robots might do and what the technology might become, but there were no robots at work in the world," he said. "It propelled new technology from ideas to implementation."

Since then, Dr. Whittaker has developed robots used inside mines and in exploration of seas and space. His team also developed the world champion robotic car. As part of Astrobotic Technology Inc., he hopes to win the $20 million Google Lunar X Prize by landing a robot on the moon in December 2010.

TMI occurred at a time when technology to clean up nuclear accidents didn't exist and few thought a cleanup would ever be necessary. Thirty years later, the experience still serves as the nuclear power industry's ultimate lesson in the importance of safety and the need to anticipate solutions to unforeseen problems.

"The perspective is, 'It will never happen again,' and it does," Dr. Whittaker said. "Then the perspective is, 'We'll be ready for it,' and we're not."