The primary objective of the event will be to provide the final 5% of the testing of the autonomous navigation system, known to loyal readers as AutoNav. To conduct the encounter, new AutoNav features need to be on board, so on Friday, June 4, the team began the lengthy process of sending new software to DS1.

The software running in DS1's main computer is three quarters of a million lines long. It took nearly 3 days for the Deep Space Network to send the more than 4 megabytes of computer code to the spacecraft. Once the software and all the supporting files were on board, the spacecraft was reconfigured to prepare it for rebooting the computer.

To run the new software required stopping the computer and restarting it. The spacecraft responds to restarting of the computer as if it were caused by a problem, so it automatically follows routines to place the spacecraft in a safe condition by turning off nonessential devices and pointing to the Sun, the only easily recognized target from its vantage point in the solar system.

The operations team knew exactly what the spacecraft was supposed to do, so prior to rebooting the computer, commands were sent to ease the spacecraft's job by pointing it at the Sun and making other changes.

On Tuesday the computer was stopped and started, and the new software has been running smoothly ever since.

As Deep Space 1 approaches 1992 KD, AutoNav will fine tune the spacecraft's path. Until 2 days before the closest approach to the asteroid, AutoNav will use the extremely efficient ion propulsion system for these course corrections.

But in the final 2 days, when time is of the essence, it will use the faster but less fuel-efficient reaction control system, or RCS. This system burns conventional rocket propellant, known as hydrazine, through combinations of 8 small thrusters.

Normally used to control only the orientation of the spacecraft, the RCS can also change the flight path using 4 of the thrusters that together will provide about 60 times as much thrust as, but use about 15 times more propellant than, the ion engine.

For small maneuvers, that propellant cost is worth it, whereas for the large maneuvers, the mass of the hydrazine the RCS would consume would exceed what the rocket that launched DS1 could have carried into space. The gentle and efficient ion engine easily wins when we can be patient.

Prior to this week, the RCS, had never been used to change DS1's course. So a test was conducted in which the RCS changed the spacecraft speed by 0.54 meters/second, or a little over 1 mile/hour. The RCS took just under 2 minutes and 130 grams of hydrazine to make this change. The ion propulsion system could have done it with only 9 grams of xenon, but it would have taken 2 hours. The successful test cleared the way for future maneuvers with the RCS.

On Monday, June 14, AutoNav made its first complete course correction. In the past it has corrected its course by modifying the direction and duration of thrusting that was planned for the ion propulsion system.

But in this case, there is no reference plan for it to change; it has to start from scratch and decide the direction and duration of thrusting. Based on its own determination of where it is in the solar system, where 1992 KD is, and what both of their motions are, AutoNav calculates what changes to make in its course.

Moreover, if the needed thrusting is in an orientation in which the spacecraft is not allowed to thrust (such as with the camera pointed near the Sun), the on-board system will figure out how to do two separate ion engine firings in acceptable directions that combine to produce the desired effect. (Our term for this process is "vectorizing".) So on Monday AutoNav will fire up the ion propulsion system to carry out its changes to assure that Deep Space 1 continues to close in on the asteroid.