View Single Post
  #14  
Old December 11th 13, 01:27 AM posted to rec.aviation.soaring
son_of_flubber
external usenet poster
 
Posts: 1,550
Default "Do It Yourself" airborne proximity warning device

This project lends "Blue Screen of Death" an entirely new dimension of meaning http://en.wikipedia.org/wiki/Blue_Screen_of_Death

I'm an enthusiastic supporter and participant in the whole Open Software/Hardware and Maker movement, but given the stakes, I have very mixed feelings about this project. A collision warning system that you cannot trust can potentially cause more harm than good. What if the system tells you to "turn right" when you should "turn left"?

A collision warning system needs to be about as reliable and trustworthy as the software and hardware that implements the ABS in your car. Implementing an ABS system properly is a whole order of difficulty beyond developing something like the very wonderful and amazing XCSoar. A defect in XCSoar is unlikely to kill you.

One thing that can in some cases make open source software and hardware more reliable and trustworthy than proprietary systems is the effect of having a large developer community scrutinizing the source code and hardware. Linar's project presently has one developer. It also helps to have a large and diverse group of users banging on the system and reporting defects.

It is an exciting time when individuals like Linar can pull together a bunch of existing building blocks and rapidly prototype a new idea. But there is an inherent fragility in any system that relies on multiple "black boxes" and a collision warning system needs to be extremely robust. Black boxes often do not do what they are advertised to do, when you use them in unanticipated ways. Just because something works when you turn it on does not mean that it is reliable.

If Linar's approach has a real potential to deliver a collision warning system at lower cost and/or higher reliability than PowerFlarm, then I hope that he obtains the funding that will allow him to pursue this project in earnest. An opensource collision avoidance system has the potential to be superior to PowerFlarm. PowerFlarm might be more reliable and better functioning if they made the software open source (but their business model does not allow that).