Originally aired September 26, 2018

Watch Now

The smart machine era will be the most disruptive in history. Medical treatment systems, the power grid, manufacturing lines, process control and transportation systems work today the same way they did 20 years ago. But in the next few years, clever engineers in every industry will find a way to leverage the amazing change in compute power and networking. That feat will threaten everyone and everything that does not respond.

The rise of connected computing will rewrite vendor relationships, redefine profitability, and re-imagine delivery from environment to cost to product. Today’s system designs must last decades. If you are a designer, your challenge is to look beyond today's experience into a future, the future, dominated by intelligent distributed computing…the Robot Overlords.

This exciting future requires a radically different architecture. The Internet’s “connect-to-the-cloud” design is fine for phones or computers talking to business systems. But most of those connections are really quite simple; your phone or computer talks to an application running in the cloud. There’s little communication between endpoints on the edge. Most connections can tolerate slow interactions at human speeds measured in seconds. This design just doesn’t work for the thousands or millions of devices that have to communicate at physics speeds in the IIoT.

The technology that can handle this is called a databus. A databus is data-centric like a database, except it communicates future information instead of storing and retrieving old information. Data centricity, in either case, is transformative and compelling. Just as databases today run essentially everything in the enterprise, databuses are a perfect fit to many of the “autonomy” challenges in the IIoT. They enable shared data between thousands of applications at full network bandwidth with extreme reliability. Databuses are key to the future of intelligent distributed systems.

This talk will dive deep into databus design. We will explain how a databus can scale to huge systems, span sensor to cloud, provide ultra-reliable operation, communicate at top speeds and adapt to changing environments. We will explore unique implementation features like serverless execution. Finally, we will illustrate the design principles with real examples in giant hydropower dams, hospital connected medical devices and NASA’s critical launch control system.


Dr. Stan Schneider - CEO, RTI