Thanks zizzle for the comments.
I agree that full automation can be tricky with due to the variations in hardware that people utilize. Because of this, I am planning to offer a wide range of options but on a select style of hardware. For instance, with temperature sensing, we generally have thermocouples, NTC's, and digital interfaces. I will cater for all three of these, but though a specific thermocouple driver, but you can use any thermocouple you want within the specs (type-k as a requirement, for instance). For NTC's, as long as the ambient temp resistance is a particular value you can use any actual model you want. For the digital stuff, I plan to use SPI/TWI, and as such once the actual data retrieval is done, I will offer both a cheap and versatile digital temp sensor, but also the option that if someone wants to use a special one, they can contact me and i'd be happy to create a new retrieval sequence. For most new sensors, once the SPI/TWI interface is setup, its just a matter of when the data comes through (at what clock cycle) and this only takes a few minutes to read in the datasheet and implement.
I have already implemented touch screen functionality and flashy graphics are on the way, with computer-interfaced plus a "computerless" system which relies on a nice big 4-line text lcd which I have already factored into cost (but will be available optionally to give more flexibility to the user)
I agree there is no way to get the firmware right for all combinations which is why I'm currently planning to make a system which relies on a slightly higher level implementation of code, such that basic firmware only handles direct hardware access and the rest is trivial for the average user to implement. Also I will be providing a "wizard" style GUI for "simple mode users" and also access to the full host of options for advanced users. User updating of the system has been a big consideration for me, and have already designed the project plan to include the cheapest, most flexible way of updating software, which users can do over the internet by themselves, or for a very minimal fee can send back a small external device to update the system (the MCU unit will actually fetch its code from a cheap SD card which is easy to remove and mail back,and also provides recepie storage for the user, which is transferrable to other units should a user go to a mate's place with the same setup).
I plan to provide excellent documentation, even instructional videos on how to hook simple things up, and for those that want to step outside the hardware options sold on my site, must fend for themselves (though im happy to help in my spare time).
The bottom line about your comment about electrocution is that this device will be somewhere between a full product and a kit. Regardless of where you go in the realm of brewing controls, you find that there is an element of high/mains voltage, and this is the users responsibility to ensure is wired up correctly.
I don't expect this to make me rich, and I understand brewers are fairly efficient/tight with thier money. However I also have learned very quickly that spending what seems a large amount of money on an item that will do exactly everything I need is often cheaper than trying to rig up a second grade solution. Just as others have tried to find replacement March pumps, the march has always won, and im sure that the vast majority of march pump owners have said that dispite the price tag it has been a danm good investment. I wish to market my device (if it actually makes it to market) in the same category. Very high quality and service, with a respectable price tag.
The motto for my system is "the affordable rolls royce of brew controllers". Even if the base unit comes out at about $200, the features available, the labour time saved and the flexibility of the system, plus the quality of the workmanship, I believe, would influence serious home brewers into purchasing one. I have also considered making a "light" version of the unit which is essentially as cheap as possible but nowhere near as many features, if people want to get their feet wet on the lower-end of the financial scale.
Also I am highly supportive of you tinkering with your own solution - it is the same reason i am tinkering with mine, and for those that want to make their own systems, I am totally welcoming of that.
Bottom line is, I will be making one for myself which will double as a market prototype, then I will ask the market if they are interested and such.
Once again thanks for the comments, I didn't find them negative - I found them realistic, which is of utmost importance in considering the success of a marketable product.
