I've looked into a few avionics possibilities too.
Certication is not the biggest problem. The existing players always pretend it is, to justify their silly pricing. For example the manufacturing cost of a GNS430 is about $300.
The biggest barrier to entry in GA avionics is the distribution/installation pipeline. The installers are very loyal to the existing principals who give them nice trade discounts and provide then with good tech support. Most avionics installers (UK and USA and elsewhere) are basically just wiremen, and while a lot of avionics installs can be done almost wholly just by reference to the wiring diagrams in the back of the Installation Manuals, many can't, and a lot of planes have weird wiring and other existing issues which one has to work with.
And one cannot bypass the resellers because the installations cannot be legally done by aircraft owners.
Obviously some things will be easier to certify than others. If you wanted to knock up a new autopilot (computer, and servos) you will have a big job, but you will have a bigger job producing the STCs for enough aircraft types to create a big enough market. There is a lot of testing involved (and not all GA AP makers have done it
). But something simple and standalone like a panel mounted CO detector would be a lot easier.
There are more openings in the non-CofA market, because you don't need to sell via installers, but that market is well covered in most respects. Look at the US homebuilt avionics market for example.
The obvious market to go for is portable stuff. That needs no approvals and can be sold mail order direct to aircraft owners. You just need to find something people want to buy
