NTSB chair Mark Rosenker says a preliminary look at video provided by the Army Corps of Engineers has shown that the southern end of the bridge behaved differently than the rest of the structure.
But Rosenker says that doesn't necessarily mean that the south end of the bridge was the source of the collapse.
"Actually, a failure at the northern end could potentially transfer loads to the southern end, where in fact the collapse could begin," said Rosenker. "So once again, I do not want you to jump to any conclusions that it's all at the southern end."
Rosenker says there is still much more evidence that needs to be gathered. He says NTSB investigators are looking at inspection reports that documented problems in the bridge structure.
They will bring four video clips back to Washington D.C., where they can enhance the images and examine them frame by frame.
Investigators will also study the bridge debris and try to fit some of it back together.
Rosenker says they will use data from these sources for what's called a "finite element analysis." It's a way to make a digital model of the bridge. They use that to run through "failure scenarios" that will help investigators pinpoint the cause of the collapse.
"We're going to start the first of a series -- a very long, comprehensive, drawn-out series -- of failure scenarios," Rosenker. "It could be a thousand of them that might be necessary until we come to the right one. It could be five."
Rosenker says those simulations will begin Monday. He says the computer program and the video provide what are tantamount to a cockpit flight data recorder in a plane crash investigation. But he cautioned that it may still take close to a year to finish their report.
The catastrophic failure of the I-35W bridge has thrown a critical spotlight on bridge safety.
Rosenker says his agency is not responsible for inspecting bridges. But he says the work of the NTSB may cause the authorities responsible for bridge maintenance to stiffen their standards.
"As we continue through this investigation, if we find regulations in any way that do not appear to be rigorous, that do not appear to be appropriate to make sure that bridges do not fall down, we will immediately make a recommendation."
Finally, Rosenker says that there are many things that could have caused the southern part of the bridge to fall in a different way than the rest of the bridge. But he refused to name them, because he says he doesn't want those scenarios to wind up in the next day's headlines.