Canadarm2 makes a V-shape in this view from a previous space shuttle crew. Photo: NASA TV/Spaceflight Now

"This flight is generally billed as an EVA flight, which makes sense given the four hard EVAs coming up," Ochoa said. "But I tend to think of it as a robotics flight. it really is a test of the station robotic arm. It's only the second time we're using the arm to lift a large payload out of the bay and this one, S0 truss, weighs twice what the airlock weighed. So it will be quite a bit different task than moving the airlock around."

At various times during the spacewalks, she said, "we put the arm in all different kinds of configurations, wrapped around the zenith side and the nadir side of the lab and often with really tight clearances. So it's really a team effort to carry out all the robotics tasks."

Ochoa will be assisted by Bursch and Walz, both of whom have extensive experience operating the arm in orbit.

"The station robotic arm is a real challenge to operate and Ellen is actually doing it from inside the lab," Bloomfield said. "There are no windows in the lab that allow Ellen to look outside and see where the arm is moving. So she's very dependent on all the camera views we can give her so that she does not take the arm and run it into anything up there.

"Those camera views come from Steve's hard work. He has put together an incredible video system that's going to allow Ellen to see all the different views and make sure that she has the clearance to safely operate the arm."

Each of the seven joints in Canadarm2 is equipped with two redundant avionics "strings" used to remotely control joint drive motors, brakes and various other systems. One string is arbitrarily considered "prime" and the other "backup."

During tests in early March, engineers discovered that the wrist-roll joint in the primary string was not working properly, apparently because of a problem with the internal electronics in that string. The backup string has no such problems and the joint worked normally. But should a similar glitch occur during S0 installation, the arm's software would consider the joint failed and put the crane into so-called "safe mode."

To prevent that possibility, engineers redesigned the trajectory the arm will follow during S0 installation to minimize use of the wrist-roll joint. In addition, Canadian programmers wrote a software patch to prevent the arm from going into safe mode if the backup string does, in fact, suffer a wrist-roll avionics glitch.

In that case, Smith and Walheim would use a power tool to manually rotate the stalled wrist-roll joint into the required orientation.

Ochoa said the crew has trained with the new procedures and that she does not expect any problems.

"While all the joint angles for the positions are different (in the revised trajectory), a lot of what we do with the hand controller, all the manual flying, is essentially the same," Ochoa said. "The main change is we've changed sort of the middle part of the maneuver, which used to be an auto maneuver that was going to move S0 forward and around the corner, getting it kind of set up for the install, which is flown manually.

"We've taken that out and replaced that with a series of single joint maneuvers," she said. "And the whole goal of that was to try to reduce moving the wrist roll joint whenever we could, to move it as little as possible, and also to always be in a position where if the good string failed and we had to go over to the string where we can't use the wrist roll, there's only one time where the EVA crew members would have to come out and actually move that joint. We could still go ahead and manually install S0 on the lab with the arm."

A new wrist roll joint with two sets of presumably pristine avionics equipment, will be installed during the next shuttle mission. Launch originally was planned for early May, but the flight was delayed to May 31 to provide more time to build a payload bay mounting fixture and to improve the angle between the sun and the station's orbit.

CONTINUE TO NEXT PAGE