I have a thermostat from the 90s with a pretty clean UI using old fashioned LCD . It allows me to set times and temperatures for "sleep", "wake", "leave", and "return" for weekdays and weekends. No silly abbreviations other than the three-letter names for the days of the week.
If I tweak the temperature it keeps it until the end of the window unless I press the "hold temp" toggle button, which locks it indefinitely.
It doesn't need a full OS and a full pixel display. Just clear, thoughtful design.
Making one of these things nice (even with a good LCD!) should really not cost much. Honestly wondering if it's worth just going into the market for making "dumb" home stuff that has a slick UI that people want.
The only problem I've seen with it is that the people who would actually think about buying them are similar to my family members: tech-savvy enough to want something like this, yet still wanting it to be internet connected. ("Why should I not be able to WiFi control my thermostat? WiFi chips are so cheap nowadays!")
The problem I see is that (a) most people don't really care and will stick with whatever they have, while (b) the people that could care usually want IoT devices and (c) the people that do care are a very small, remaining portion of the rest (e.g., some proportion of the usual HN audience).
You can't get VC for any consumer hardware product that doesn't generate a continuing revenue stream. That means either charging a monthly fee for cloud connectivity or selling customers' data. Both require Internet connectivity, even though that doesn't really add much objective value (and creates huge security problems) for so-called "smart" home devices.
As far as I can tell, you wouldn't need VC funding for that anyway. A thermostat, as far as I can tell, is really a set of relays that are toggled by a circuit board connected to a thermometer. A quick search on the topic brings up a lot of successful attempts at making thermostats with Raspberry Pis. It doesn't seem like advanced hardware by any stretch of the imagination.
Prototyping should be very easy. I'm sure that a Pi Zero would suffice, along with a board with relays connected to GPIO as well as a thermometer circuit, and an LCD screen. The production product could use a custom built stripped-down ARM board with integrated components, and probably be made for cheap. I'd pay $150 for a dumb thermostat that has a nice LCD screen. Even better if it's hackable(i.e. I can install my own software on it).
RPi is overkill. It could easily be done with an Arduino--including wifi if you wanted. The tough part is building a business around it. But maybe if you focus on sustainability rather than growth and exit, you could do it without VC money.
If I tweak the temperature it keeps it until the end of the window unless I press the "hold temp" toggle button, which locks it indefinitely.
It doesn't need a full OS and a full pixel display. Just clear, thoughtful design.