DataWeek Conference + Expo

With affordable Micro-Electro-Mechanical Systems (MEMS) manufacturing economics, and the networking protocols solving the kinks to connect low-powered devices, we are on the verge of unleashing the promise of ubiquitous computing with the Internet of Things. Data-driven applications built for these new use cases to provide context and awareness will have their own unique constraints. Obviously, the large volumes of data accumulated for analysis by sensors demand a “big data” infrastructure. However, other requirements exist. With the proliferation of “Things” and the desire to provide valuable insights by correlating their readings, data-driven applications will become very complex. From elementary applications of today that provide dashboards from readings of embedded devices, the next-generation devices will be aggregating, analyzing data from multiple families of Things, providing predictive and other novel analysis. In addition, the speed to respond — or response latency — will drive the efficacy of the applications. While the applications will have to adapt to the optimal technologies of today, they must also be prepared to quickly leverage new innovations in analytics as and when they come.