I've had similar questions from some of my hard core embedded colleagues, here's my answer: Just think of something you might do with a traditional embedded sy

I've had similar questions from some of my hard core embedded colleagues, here's my answer: Just think of something you might do with a traditional embedded systems approach, then google "arduino for ________" or "arduino shield for _______" and see what you'll find. Its can be pretty amazing. Its much broader that the surface stuff of LEDs and basic sensor / actuator tasks. As someone who's done production embedded systems work, what I build from arduino (and the fairly large arduino ecosystem) is not for high volume product, its for proof of concept, small run pilots and show and tell (and fun). Off topic a little: One thing I've done to convince colleagues of the a plausible story of arduino POC to Production is collect a bunch of arduino gear (simply through web search and amazon) and build a paper-product with all the cobbled together parts. Then pose the question "if these cobbled together things cost _______ in single quantities, is there a plausible story for them to be 10x (or 100x) less on a cost optimized board with a proper contract mfg"? If the answer is yes, then you cobble together your arduino thing and go show and tell with customers.