Why The Ariane 5 Failure

The Ariane 5 accident is incredibly interesting in terms of what failed, how, and why. Essentially, the Ariane 5 is a European rocket that, thirty-nine seconds after take-off, veered to the side and started disintegrating in the atmosphere. After extensive review, the problem came from a conversion error in the code for the rocket’s inertial reference system, from a 64-bit floating point number to a 16-bit integer (Lions, 1996).
Now, it’s not shocking that there was an error in code for a rocket: there are, of course, going to be thousands of those, and they get caught habitually. What was concerning is that the tests did not catch this error. The reason the error made it in, at all, was that this particular system had worked before on a different rocket, the Ariane 4. However, the Ariane 4 was balanced differently, and so the problem of physics that gave rise to the system error didn’t occur.

The error managed to make its way into the software, but it’s a case study in software engineering for a reason: it appears that there are several different places that could be the root of the problem. There was a failure to test the system with the actual inputs the Ariane 5 would be putting out. There was a failure to give clear reasons as to why some conversions were protected and not others. There was a failure to include the measurements of the Ariane 5 in the specifications. There was a failure to re-examine the systems put in place for the Ariane 4 vs the 5, including both measurements and length of process. There was a failure to communicate across departments and branches, to write software that assumed software is fallible, and a failure to document reasons for doing any of these things (Lions,