Unless, that is, NASA and its project partners can come up with a quick fix for a serious problem with the stations new $600 million Canadian robot arm, an all-important device required for raising station segments that cant be mounted with shorter shuttle cranes.

"If we cant fix the station arm, then were going to have to stop and get it fixed before we start installing more modules, certainly on the U.S. side at least," NASA flight director John Curry told reporters in a recent news briefing.

"It is literally in the critical path for many, many flights that weve got up and coming in the future, because as we make this station bigger and bigger and bigger, you just cant reach [construction sites] with the [shuttle] arm."

Two and a half years after the stations first building blocks were launched and linked in space, NASA and its 15 global partners now are wrestling with the most significant technical glitch to hit the $60 billion construction program.

The shoulder on the stations newly installed robot arm is acting arthritic at best, and engineers in both the U.S. and Canada are scrambling to come up with a remedy that will negate or at least stall -- the need for joint replacement surgery.

Ground engineers now face an end-of-June deadline to come up with a fix. The other option: Putting off all future U.S. station assembly flights until a shoulder repair mission can be conducted.

The latter alternative essentially would push back launch of the next U.S. station component an airlock that will be used to stage spacewalks at the outpost until September at the earliest.

In that scenario, however, the airlock installation mission could end up facing a delay until late this year or early in 2002, a major setback that project managers and the stations current resident crew are trying to avoid.

"This is obviously something that needs some attention, and [ground engineers] have been working, I think, around the clock coming up with solutions," station flight engineer Susan Helms told reporters in a space-to-ground news conference earlier this month.

"And my reaction, of course, is that I hope that the workarounds and the questions that they have get resolved so that we can continue to use [the arm]."

Delivered to the outpost in April, the new construction crane is an advanced version of the 50-foot (15-meter) Canadian robot arms carried aboard space shuttles.

Considered the heart of Canadas $900 million station contribution, the 57.7-foot (17.5-meter) station crane is capable of crawling to places that cant be reached by the shuttles fixed robot arm, which has been used to perform all outpost construction to date.

Snare-like hands on either end of the seven-jointed station arm are designed to grapple fixtures that electrically connect the crane to computer control stations inside the complex. That capability is meant to enable the arm to move end over end, from one pin-like anchor to the next, "inchworming" to different work sites outside the outpost.

Equipped with force sensors that give it an artificial sense of feel, the crane is capable of lifting four times more weight than its shuttle predecessor. And since it will have to operate in orbit during the next two decades, the arm can be retrofitted with spare joints if necessary.

In order to operate the crane, computer commands are sent from a robotic workstation inside the outpost to a computer control box located near the elbow of the arm. Those commands then are relayed to an electronics unit on the arm joint that operators are attempting to move.

Like most space station components, the arm is equipped with redundant systems. Prime and backup robotic workstations now are located inside the stations U.S. Destiny laboratory. Prime and backup computer control boxes are located near its elbow. And its joints are equipped with prime and backup electronics units.

NASA calls these integrated sets of equipment the "prime and redundant strings" of the robot arm, a reference to the electronic pathways over which arm computer commands travel.

To date, the prime string has been working as advertised, and in that mode, the construction crane can be operated as intended. But in recent weeks, the shoulder joint on the redundant string has not been working.

Engineers think the problem might be a faulty backup electronics unit on the arms shoulder joint.

And if any joint fails to work properly, then the arm puts itself into what NASA calls a "safe mode" which means that it automatically stops operating.

Therein lies the heart of the dilemma.

NASA flight rules call for both the prime and redundant strings to be fully functional before the arm can be used to install the $164 million airlock, which now is scheduled to be hauled up to the station aboard shuttle Atlantis in early July.

The flight rule is intended to avoid a scenario during which failures on both the prime and redundant strings could leave the 12-ton airlock stranded at the end of the crane while astronauts are trying to mount it to the station.

"Thats ultimately the concern," said Curry. "In the worst case, if you lost the prime string and had to switch over [to the redundant string], then you dont want to get stuck."

Shaped like a giant genie bottle, the airlock is to be plucked from the shuttles cargo bay with the station arm and then mounted to a berthing port on the starboard side of the U.S. Unity module, which serves as a pressurized passageway to all parts of the outpost.

A robot arm failure midway through that job could leave the airlock stuck in a position that blocks the approach path to station docking ports.

Whats more, once the airlock is removed from the shuttles cargo bay, it must be hooked up to station power supplies within hours to avoid damaging its thermally sensitive systems. The concern here is that station power cables might remain out of reach if the arm shut down during the installation job.

The shuttles robot arm, meanwhile, couldnt be pressed into service because it isnt long enough to do the airlock installation work.

Station project officials, consequently, are hesitant to embark on such a high-stakes job unless the redundant string on the arm is operating well enough to perform the work if prime units for some reason fail.

With the near-term future of the construction project hanging in the balance, American and Canadian engineers are in the midst of hatching a recovery plan that would enable NASA to proceed with the airlock installation job despite the arm problem.

To that end, a diagnostic computer software program will be run this week in an effort to pinpoint the root cause of the shoulder joint problem.

Then later this month, a "software patch" will be beamed up to the station to essentially mask the problem encountered with the shoulder joint. The idea is to enable the redundant string to continue operating even if a shoulder joint failure crops up during airlock installation.

As a backup to that plan, NASA also is devising a way for spacewalkers to manually move the shoulder, if need be, swinging the airlock into a position where it could at least be hooked up to station power cables if not mounted to the outpost.

But if neither of those plans pans out by the end of June, then NASA likely will delay the airlock installation flight and press ahead with a mission to remove and replace the suspect shoulder joint.

A new joint now is being readied just in case, and the station's current crew stowed the arm Tuesday in what NASA calls a "joint-maintenance position." The idea: to make certain that the arm can be cradled properly for any shoulder-joint replacement surgery.

That job if it becomes a necessity -- could fall to astronauts scheduled for launch onboard shuttle Discovery in early August. Their prime mission now: to ferry a new resident crew up to the station and then return to Earth with Helms, fellow U.S. flight engineer Jim Voss and outpost commander Yuri Usachev.

An even gloomier scenario: delivering a new shoulder joint in August but deferring shoulder surgery until a shuttle mission now slated for launch in late November. And in that case, the airlock installation mission likely would be grounded until early 2002.

Station project managers, however, are confident that their current recovery plan will enable the airlock installation mission to be launched without major delays.

"Im hopeful that we can come up with an option thats going to allow us to have the redundancy to go ahead and install the airlock in the sequence that we want," said Curry. "Thats our goal."

And while Curry and others say the shoulder problem ranks with the most significant project challenges to date, they also say that more of the same should be expected during the course of what many consider the most complicated civil engineering endeavor in history.

"Im not at all surprised that weve run into some things like this," Voss said. "The space station is an extremely complex machine. It combines so many different things together, so many different computers and systems, pieces from other countries. Its remarkable that it works at all."

"To be honest, Im amazed and pleasantly surprised that weve gotten this far with as few problems as weve had," added Curry. Its a challenge, though. And it always will be so."