Updated 8:30 a.m. May 30: Software patch did not work this morning; possibility of joint replacement and shuttle shuffle grows
Updated 10:15 a.m. May 30: NASA to decide shuttle schedule June 11
Updated 11:00 a.m. May 30: Atlantis officially bumped to July
Updated 11:30 a.m. May 30: Adding quotes from ground controllers on today's attempted software fix

Shuttle Atlantis sits inside the Vehicle Assembly Building on Tuesday night after its quarter-mile trip from the Orbiter Processing Facility hangar. Atlantis will be attached to its external fuel tank and solid rocket boosters in the VAB. Photo: NASA OTV

Problems with the international space station's new robotic arm has forced NASA to delay the next space shuttle flight, and the possibility is growing that a daring repair mission might be needed to replace one of the crane's joints before construction of the outpost can continue.

The $600 million Canadarm2 was delivered to the station last month, becoming an integral part of future assembly chores for the orbiting complex.

However, recent testing of the 57-foot long crane has uncovered glitches that have the potential of scuttling NASA's summer shuttle launch schedule by flip-flopping of the next two missions if one of the arm's joints has to be swapped out.

The arm must be working properly in order to accomplish the next station assembly task -- hoisting the $164 million Joint Airlock from the payload bay of shuttle Atlantis and mounting it to the Unity node. The shuttle's robot arm can't perform airlock installation because its reach isn't long enough, which means the continued construction of the station relies upon Canadarm2.

Atlantis had been targeted for liftoff on June 20, some six days later than originally planned due to extra work needed to dry 600 rain-soaked tiles on the shuttle's belly.

But senior managers today decided to bump the launch until the first week of July to allow more time to determine how to fix the arm, and the flight ultimately could be delayed to September if the NASA decides to mount a repair mission to the station before the airlock flies. A decision is expected on June 11.

Atlantis moved out of its hangar and into the cavernous Vehicle Assembly Building at Kennedy Space Center on Tuesday night. The ship will be attached to its fuel tank and twin solid rocket boosters inside the 52-story VAB this week, but rollout to the oceanfront pad won't occur until NASA decides on a launch date.

The arm troubles began about two weeks ago when routine testing -- being performed by the Expedition Two astronauts -- was halted when the crane suddenly entered a "safe mode" and its brakes engaged while the Wrist Roll joint was being moved. The brakes essentially freeze the arm in place.

Subsequent tests have not been able to repeat the incident, but engineers suspected the Arm Controller Unit was to blame. The ACU is a computer mounted on the arm, near its elbow.

The problem occurred when the arm was being operated through the redundant control string. This "string" is the pathway that computer commands follow from the ACU to the joint, telling it what to do.

The international space station's new robotic arm, seen here forming a V-shape, has experienced two problems in recent weeks. Photo: NASA

Since there is a spare computer unit aboard the station, officials have studied the option of having station residents Yuri Usachev and Jim Voss swap out the ACU during an already-scheduled spacewalk on June 8.

The job would be added to the spacewalkers' "to do list" during the station's first spacewalk. The excursion's original goal is to reposition a cone-shaped docking mechanism on the Zvezda service module in advance of the next Russian segment -- the Docking Compartment 1 -- arriving later this summer.

As the brake problem was being examined, a separate and more serious failure arose with the arm's Shoulder Pitch joint. The joint's electronics are unable to receive commands from the ACU via the arm's backup control string.

When the joint failure manifests itself, the arm's entire redundant electronics string is paralyzed, rendering the crane's backup control side useless. Without the redundant path, the arm would have to operate only on its primary side with no margin for failure during operations.

To date, the arm's primary command string has worked normally. However, embarking on a high-stakes job such as installing the airlock with no backup systems on Canadarm2 is something NASA wants to avoid.

NASA officials are uncomfortable having the arm install the airlock without its redundant side healthy and ready to take over if the primary string develops a glitch.

Analysis of this second problem indicated that the glitch could be solved by a software patch, which was uplinked to the station early this morning.

But the patch didn't cure the problem, leading some engineers to suspect there is a real problem with joint itself.

"We uploaded the software patch for the arm and it did the things the patch was designed to do. But it did not, as a by-product, fix the problem we are seeing between the Shoulder Pitch and ACU," astronaut Dan Burbank radioed station resident Susan Helms this morning. "We are seeing that same signature with loss of (communications) between those two ORUs (Orbital Replacement Units)....obviously that has some pretty big implications.

"There are still some more options out there but the prospects are not looking very encouraging right now and it seeming to indicate we may have a problem with the the Shoulder Pitch joint that would require an R&R (removal and replacement).

"There are still some other long-shots out there, some software patches that we are looking at right now that might cure the problem. But the most likely case seems, at this point, to be a hardware problem with the Shoulder Pitch joint."

Animation shows Canadarm2 hoisting the airlock from the shuttle and mounting it to the international space station. Photos: NASA

If engineers can't solve the joint problem with software or develop a way to work around the glitch in some manner, the only other option is replacement of the Shoulder Pitch joint with a spare in a dramatic spacewalk by shuttle astronauts.

Such a transplant would call for the STS-105 mission of shuttle Discovery -- originally slated set for launch July 12 -- to fly before Atlantis' flight to deliver the airlock.

The need for additional training time for the astronauts would bump the Discovery flight into August, which is also after an orbital mechanics blackout period extending from the second-half of July into the first week of August that prevents the shuttle from launching to the station.

Discovery's flight, which already had two spacewalks booked on the agenda for crewmembers Dan Barry and Patrick Forrester, would also exchange the station's crew by ferrying Expedition Three to the outpost and returning Expedition Two to Earth.

With the arm repaired, Atlantis would be launched approximately a month later with the airlock that will serve as the gateway for future station spacewalks by both American and Russian astronauts.

But before the airlock flies into space, engineers must unravel and correct the mysterious cause or causes of the arm ailments.