Engineering Internet of Things systems
The Internet of Things (IoT) has become an academic discipline, with some universities offering degrees in the Internet of Things. Such a rapid progression should come as no surprise, for the Internet of Things is the most significant technological trend in engineering, and it brings a new way of thinking.
Traditionally, engineers have been specialists either in software or in hardware, having been taught in college to design for a disconnected world in which every system, no matter how complex, had clear boundaries. Whether designing products, plants, heavy equipment or buildings, engineers have become accustomed to known requirements and more or less static environments.
The age of data, sensors and connectivity has introduced another paradigm, bringing challenges, expectations and possibilities never before encountered. But what differentiates the design of IoT systems from the design of traditional complex systems?
Machine sensing and feedback loops, which have long been integral parts of control theory, have become possible on an enormous scale through the connection of low-cost sensors to cloud-based platforms providing analytics and security. The availability of operational and maintenance data that results has changed conventional wisdom about engineering practices and tools.
Data has transformed industries such as retail, banking and insurance, giving rise to concepts such as business intelligence, the single customer view, multichannel marketing and financial market technical analysis. These and other such concepts have been made possible by the ability to monitor, analyze and react to business data generated by millions of transactions.
In much the same way, data is poised to transform engineering. In an open connected system, data sharing between parts of the system is infinitely flexible, restricted only by design. But such an approach requires a paradigm shift on the part of engineers from a closed world to an open one.
When designing a complex, yet disconnected product, engineers clearly understand the boundaries of the system and design all connections between components. By contrast, engineers who design for a connected world must allow for infinite connections inside and outside the system, deliberately restricting connections to achieve desired outcomes and for reasons of security.
Such changes in the basic assumptions of the design process have reshaped fundamental engineering practices. Development of new tools will be integral to the design of secure IoT systems, allowing engineers to base their decisions on reservoirs of data sourced from the field, from manufacturing or even from competitors. Moreover, foundational questions begin to drive the design of systems that can sense usage, performance, system status and other data well beyond their own boundaries: What data matters to you? What key performance indicators are important to your system? How do you secure your system?
Equipped with the right monitors, a physical system can shorten reaction times by raising alerts. For example, real-time monitoring of temperature, flow and pressure can provide data integral to maintenance of a water management system. Analyzed independently, such values are “just” streams of data, but adding engineering knowledge to the mix allows identification of water leaks, malfunctioning valves and more, enabling engineers to respond quickly, deliver changes on the fly, test assumptions and implement a closed design–build–test–deploy–sense–react cycle.
Engineering data is not only data on a big scale, but it is also heterogeneous data. Such data originates from a variety of sectors, among them systems, software, mechanical, electrical, manufacturing, maintenance, facility management, testing and operations, and must be integrated into a consistent and meaningful stream of engineering information as a prerequisite to industry transformation. Accordingly, the engineering industry is adopting and developing standards to facilitate integration and encourage cross-references between engineering data elements.
Engineering technologies in the age of the Internet of Things, such as continuous engineering, create connections to IoT operational data and provide tools and standards that can be used to connect design, testing and operation to sophisticated analytics to enable a closed-loop engineering process for the Internet of Things.
For example, suppose an automotive engine manufacturer that uses IBM technology relies on data collected from its deployed system to tune and calibrate its design environment. The manufacturer tunes engine simulation models by correlating performance data about speed, temperature, vibration, noise and more with environmental data about temperature, humidity and geolocation—it doesn’t stop there. During data collection and analysis, the manufacturer updates the simulation parameters, enhancing the simulation and boosting the accuracy and reliability of verification.
To connect with other engineers who are immersed in designing for the Internet of Things, attend my session on designing for the Internet of Things at the IBM Continuous Engineering for the Internet of Things (ICE IoT) conference. In addition to technical sessions, the conference will offer opportunities for skills development and will feature prestigious keynote speakers hailing from a variety of engineering fields. Discounted tickets are available for some engineering organizations; for more information, please inquire in the comments section.
If you see me in the hallways at ICE IoT, stop and say hello. I look forward to seeing you there!