New Horizons was launched on 19th January 2006. It has been travelling for 3463 days (9.5 years).
In 2006 we were still using Windows XP and Office 2003. In 2006 I wasn’t even working in IT. As a software developer it is hard to imagine writing software that won’t be used for another 9.5 years.
The spacecraft carries two computer systems: the Command and Data Handling system and the Guidance and Control processor. Each of the two systems is duplicated for redundancy, for a total of four computers. The processor used for its flight computers is the Mongoose-V, a 12 MHz radiation-hardened version of the MIPS R3000 CPU. Multiple clocks and timing routines are implemented in hardware and software to help prevent faults and downtime. To conserve heat and mass, spacecraft and instrument electronics are housed together in IEMs (integrated electronics modules). There are two redundant IEMs. Including other functions such as instrument and radio electronics, each IEM contains 9 boards. The processor distributes operating commands to each subsystem, collects and processes instrument data, and sequences information sent back to Earth. It also runs the advanced “autonomy” algorithms that allow the spacecraft to check the status of each system and, if necessary, correct any problems, switch to backup systems or contact operators on Earth for help.
For data storage, New Horizons carries two low-power solid-state recorders (one backup) that can hold up to 8 gigabytes each. The main processor collects, compresses, reformats, sorts and stores science and housekeeping (telemetry) data on the recorder – similar to a flash memory card for a digital camera – for transmission to Earth through the telecommunications subsystem.
Communication with the spacecraft is via X band. The craft had a communication rate of 38 kbit/s at Jupiter; at Pluto’s distance, a rate of approximately 1000 bit/s. To put this speed into context, the size of the image at the top of this page is 649kb and would take over 86 minutes to travel from Pluto to Earth, (I am sure the actual image was much larger and took even longer to reach us.) Besides the low bandwidth, Pluto’s distance also causes a latency of about 4.5 hours (one-way). The 70 m (230 ft) NASA Deep Space Network (DSN) dishes are used to relay commands once it is beyond Jupiter. It will take 16 months to send the full set of Pluto encounter science data back to Earth.
Most software encounters a problem from time to time, but the software on board this spacecraft has to keep running for almost a decade. Due to the distances involved its not possible to upload any code tweaks. Any commands that are sent to the spacecraft will take 4.5 hours to even get there so there needs to be a huge amount of autonomy and redundancy to cope with such a massive journey.
I can’t even imagine how you would begin to write the software to run such a craft. I don’t think I have written any code that is as old as is running on the New Horizons craft, and if I had I would guess the quality would be quite poor certainly not good enough to work non stop for a decade.
There are some incredibly clever people out there who have been involved in this project. Congratulations team you have every right on being proud as the data starts to trickle back to earth.
This post has already been read 1151 times!