I'm eager for IoT, except that so far, most IoTish things have been rather grotesquely proprietary and ad-hoc. if I get a d-link camera, it can route AV through a d-link provided web service, and d-link even provides ios and android apps. but it's still mostly locked into d-link's network, with little in the way of API for me to use for anything else. (yes, I know how to access the camera directly on the LAN.) if I get FitBit products, they assume the presence of a corporate cloud service as well. same with wifi-enabled thermostats, etc.
the main point is that I own these devices, so their first level of functionality should be to deliver data and control to me. some kind of cloud access should be an optional add-on. after all, what reason do I have to expect the vendor to continue providing cloud access? perhaps more important than continuity is security: if each IoT device implements its own idiosyncratic cloud-access mechanism, my security exposure increases dramatically. is there any reason not to regard them all as potential trojan horses? or at the very least, incredible privacy leaks.
if the IoT industry wants to do a good job, they'll get a handle on this before it becomes any harder. there are many examples, especially related to the computer industry, of standards that provide the right levels of interface, control and interop, while avoiding the tar pit of licensing. (FRAND is a fraud - it is only unencumbered standards which succeed. mp3 is a good example: Fraunhofer attempted to impose a fairly FRAND licensing scheme, and it was a historic failure - standards want to be free.)
the opportunity is to provide an interface that is secure, local and configurable. the goal of IoT should be first to let owners locally fetch from sensors and control actuators. maybe there will even be a secondary market for devices that add smarts (analytics, sensor-decision-actuator smarts, cloud access extensions) to IoT devices.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...