Getting Started with IoT

Manufacturers can scarcely pick up a trade publication or listen to the business news these days without encountering articles about the Internet of Things (IoT) and how this paradigm-shift promises to revolutionize industry from top to bottom.  Yet many traditional manufacturers may not see the immediate usefulness of connected production machines, “smart” products, and “smart” facilities, or may be unsure of how to begin approaching the problem of modernizing even if they are excited about the opportunity.

Within the areas of remote control, remote sensing, and remote monitoring there have been tremendous advances in recent years that make designing and building customized IoT solutions approachable from a budget perspective.  In my own experience as a designer and builder of IoT devices and technologies, the most significant enabling technology for small businesses has been the rise of ultra-cheap, high performance computer systems like the Raspberry Pi, Arduino, and other System-on-a-Chip (SoC) devices.  For those who may never have encountered systems like these before, the Raspberry Pi (for example) is a tiny circuit board with capabilities comparable to a smartphone, with some models priced under $30 per unit. At these price points, it becomes possible to imagine putting many of these devices into service on the factory floor, inside products,  or in other situations where previously leveraging computing power would have been too expensive to consider. When combined with low-cost cameras, thermocouples, pressure sensors, accelerometers and other sensors, these components can create the backbone of a fully custom IoT solution at price points that can fit within the budget of many organizations.

However, having a bunch of cheap computers running in your shop or inside your products is not terribly useful if these devices cannot speak to each other and to the outside world.  A comprehensive technology platform will also need to consider how the devices or data will be used. There have been tremendous advances recently in these areas as well, from cloud computing to extend the capabilities of the individual devices, to business intelligence approaches for large scale data mining to process all that data, to machine learning and artificial intelligence to draw inferences from patterns and predict future behavior.  These technologies have matured into an increasingly convergent set of tools and techniques that can be effectively developed and deployed by smaller software development teams and applied to the problems of “ordinary” small to mid-sized businesses.

Along the way, many manufacturers and industrial companies that choose to invest in these new approaches will generate valuable intellectual property and novel technologies that may have additional value as commercial products.  Working with a local technology development partner allows for the close collaboration that tends to lead to innovation. Even without spinoff products and services, the combination of these forces, properly leveraged, can drive improvements in productivity, reduce costs and waste, and improve visibility into real-world production problems.

The first step for many manufacturers may simply be to start conversations with technology companies, industry peers, and their own front-line staff who may have ideas for how processes can be improved and where better data can drive faster, better decision-making.