Sandy Klausner: "There is a general belief that the Internet of Things (IoT) will shape a newer "information society" and "knowledge economy." However, there are fundamental technological barriers that are attenuating the transition from the present "Internet of PCs" towards an "Internet of Things" (IoT) in which 50 to 100 billion devices might be connected to the Internet by 2020. The primary reason the 'thing' ecosystem will remain largely unrealized is the lack of an intelligent network infrastructure to support it. Such an infrastructure requires new Internet software architecture and an underlying applicative design methodology to transform connected objects into real actors at scale.
The manner in which software is developed hasn't fundamentally changed since the 1960s. Character string-based programming languages have ruled computer science, but even the latest dialects such as Go and Dart are not likely to materially improve software productivity or enable a broad population to participate in the IoT systems creation process. Now after several thousand dialects, we can reasonably ascertain that legacy computer science will not produce the exponential improvements needed to manage the overwhelming interoperability, contextualization and security challenges of global scaling.
The orchestration of a paradigm shift is essential if the software industry is to ever become at least as innovative and productive as the hardware industry, which is following Moore's Law. A new software science approach needs to be established to meet the requirements for the emerging IoT of unattended devices. This 'clean slate' approach can extend processor architecture ecosystems up through the software layers in a holistic manner to greatly simplify the build-out of the IoT.
In this paper, I describe an intelligent network infrastructure (Smart Web) that uses a new computing platform (Cubicon) to support the IoT. Intelligence is achieved by embedding machine reasoning capabilities into the network itself."