The U.S. Army has run into technical problems with its 155mm Excalibur "smart shell." In recent testing, the shell performed as expected (using GPS guidance to land within 20 feet of the aiming point.) But there were times when the shell was unable to get the GPS signal, and just followed a normal (ballistic) trajectory, without the backup (and less accurate) inertial guidance system kicking in. This can be a problem with Excalibur, because the justification for the $50,000 shell is its ability to hit a target in situations where friendly troops are nearby. If the Excalibur shell does not get the GPS signal, you have to make sure it's unguided trajectory, or less accurate inertial guidance system, will take it where there are no friendly troops or civilians. Having to do this every time you use Excalibur can be complicated, time consuming, and often not possible. But these problems are deemed solvable, and introduction of Excalibur (in Iraq) will be delayed from this Summer to this Fall.
Getting "smart shells" to work effectively is nothing new. Back in the 1980s, the 155mm Copperhead round was developed, at great expense, to take out tanks with one shell. The Copperhead was laser guided. That is, it homed in on laser light that a forward observer was creating by pointing the laser at the target. It was the same technique used with laser guided bombs.
But this was expensive technology. Each of the 3,000 Copperhead shells cost several hundred thousands dollars (the price varied, up to half a million bucks, depending on was doing the calculating). While a "dumb" artillery shell will land with 75 meters of the aiming point, the Copperhead would land within a meter or two. But so what? It turned out there were many easier, and cheaper, ways to destroy enemy tanks. This was demonstrated during the 1991 Gulf War, when a few Copperhead shells were used, successfully, but to reactions of, "whatever."
Then again, Russia developed its own version of Copperhead, Krasnopol, and sold some to India. During a 1999 war with Pakistan, high in the Himalayan mountains, Krasnopol proved very useful in taking out enemy bunkers, without causing avalanches or destroying the few pathways up the steep hills. The Indians paid about $40,000 for each Krasnopol shell (two thirds what the Copperhead was supposed to cost), and found it a good investment. This encouraged the American developers of the next generation smart shell, Excalibur, which already had several years of design and research invested.
Right now, the Excalibur will be competing with the new U.S. Air Force "small diameter bomb" (SDB). This 250 pound device, which looks like a missile, but is an unguided smart bomb, weighs twice as much as Excalibur, and thus produces a bigger bang. But you need an air force bomber overhead to get a SDB, while army artillery is always there. You also need an air force FAC (Forward Air Controller) nearby to call in the bomb, while there are many more army personnel who can call for artillery. The SDB costs about as much as Excalibur. Another competitor is the GPS guided MLRS rocket. But because rockets are less accurate than artillery shells to begin with, GPS guided MLRS cannot hit targets as accurately as SDB or Excalibur, and is already in Iraq.
The third generation smart shell is also in development. This is the Projectile Guidance Kit (PGK), which is actually a large fuze, that screws into the front of a 155mm or 105mm shell. This longer fuze contains a GPS and small fins to guide the shell to a precision hit equal to an Excalibur shell. The army doesn't expect to be passing these out to the troops for another five years. But if development goes smoothly, and Excalibur proves useful and popular, then the PGK might show up earlier. The PGK will cost less than half what each Excalibur does and, more importantly, can turn any shell into a smart shell. This is important for artillerymen, who don't like to carry around a lot of special shells, just in case. Artillery units already carry several different types of fuzes for their shells, so one more is not seen as a burden.
The pinnacle of artillery operations has always been, "one shot, one kill." But achieving this has always been like a golfer getting a hole in one. It can be done, but it's rare. Smart shells make "one shot, one kill" commonplace, and means artillerymen will spend less time constantly replenishing their ammunition supplies. Firing the cannon less often is also nice, as those beasts are a bitch to keep clean.