mpp4manu

Authored Comments

The rise of 3D printings, particular those that use materials appropriate for load bearing structures, mean an open source car may be closer to reality than an airplane. In this case, assuming certain critical safety systems are assured (brakes, steering, crash protection), a user designed car may be in our future.
One of the major auto manufacturers had a "skateboard" concept where they used a common chassis and powertrain that could be mated to a variety of bodies. That concept would provide the public with the opportunity to design a body that met their practical and aesthetic requirements, while reserving the truly capital and engineering intensive aspects of the vehicle were produced to strict, tested standards.

What most people don't know is that there are extremely rigid requirements for producing any software for use in commercial aircraft. The DO-178B document details criteria that should (or must) be considered for any software used in airborne systems. Few of us, myself included, have ever labored under the level of scrutiny it imposes.

But as Bob Jones points out, there are reasons for these design criteria. Hard experience, purchased with human lives, has determined a software development model that not only assures accountability, but also provides a consistent analysis process that permits precise fault analysis.

How often have you heard of a airplane crash traced to software? The answer is, none that I can think of.

That said, NASA, the real rocket scientists and progenitors of these types of development processes, managed to crash the Mars lander into the planet because they got their measurement units mixed up.