I am thinking of founding a start-up company with a stress on applied research in some area of computer science. No any specific difference with other product-centric startup except for more research activity than usual in standard R&D process. Kind of private research laboratories. Clearly this would imply higher risk of failure and other deviation from conventional startup practice (e.g. grants as one source of funding). So I am wondering is there any experience, thoughts, resources to help me with the endeavor?
I'm not sure where you're going to be working - this answer has a UK perspective.
Universities have increased their focus on commercialising research - so a focus on applied research should make you gravitate towards building relationships with academic institutions.
Many of those have groups, departments or companies focused on facilitating that process. They're potentially your friends in this process. Applying research is possible, applying research with access to the researchers is fun, and potentially highly lucrative!
It's entirely usual that many of those discussions don't jump straight from the research to the commercial result - there's additional exploratory research on the way. It sounds like that's your skill: to take that further means being ready to spend some time raising money too.
Definitely you should explore grant funding. You can very often fund a significant part of the total costs by accessing grants. Get advice, put in applications, learn the ropes. But you may also find it useful to see what funds, angel groups etc are operating in your area and start to build links. Whether that's an immediate source of funding for you or a way of accelerating things downstream as you're getting ready for commercialisation, those links will be beneficial.
As a final point, you need to be careful to keep in mind the high risk of failure. Be clear about your strategy here. You can aim for breadth to spread the risk - keep costs low, be systematic about spotting the stars and clear about stopping projects when it's clear they aren't progressing. Or you can commit totally - go narrow and deep. In this latter case, your funders or investors know they have bought into the risk - and you know you have to go all out. I've seen both strategies work. But fuzzy compromise in this area is confusing, damaging and dilutive.
Good luck! Hope it works out for you!
Well, what's about getting in contact with some Universtity Institutions which are nearly working in the supposed area of your startup idea? Try to get some knowledge from them and see what you could do better.
I guess there would also be a difference in doing applied research for Computer Hardware or Software (or would it be both in combination?).
What will be the most important benefit for your future clients? And what will your target clients are, small or big businesses? The big computer and software companies might have great research facilities, but may not focus on your area of research. Did you have any knowledge about what they may lack?
If you do applied research you might be eligible for government grants, but this may depend on the country where you live and on the requirements of such programs.
Hope this helps you to get a bit further.