Don’t Put the IoT Cart Before the Horse: Bsquare Interview

Please describe a recent IoT project you have worked on or have observed that was substantially better, faster, smarter or more efficient than an older technology?

Bsquare has recently deployed its DataV IoT software into a heavy-duty trucking environment in order to help the truck manufacturer offer greater uptime, reduced service costs, and accelerated warranty payments. This is one of the first IoT deployments in the industry to fully “close the IoT loop” and deliver tangible benefits not only to the truck manufacturer but to their own customers as well.

This project captures real-time engine data for delivery to cloud-based data stores and IoT applications. Bsquare DataV applies machine learning and advanced data analytics to provide predictive reasoning, i.e., the ability to know something’s going to go bad before it actually does so that remediation (which data analytics speeds as well) can be handled in more orderly, less expensive fashion. DataV then goes a step further by automatically orchestrating actions necessary to pro-actively repair the truck.

The bottom line: faster, more accurate, fully automated processes that yield better asset uptime at lower cost.

What do you see as the biggest potential of the Internet of Things?

The biggest potential associated with IoT is the ability to apply advanced data analytics, holistic rule processing, and automated process orchestration in industrial settings so that real-time automation can finally be achieved. Most IoT experiments to date have largely been focused on data extraction and data visualization. While these are necessary initial steps, they do not deliver the benefits businesses are looking for in IoT initiatives.

In order to fully realize IoT’s potential, data analytics and automated orchestration of actions and processes need to be added to the IoT system. By doing so, businesses move toward complete IoT systems—data extraction and monitoring coupled with applied analytics and orchestration resulting in completely automated, highly accurate, self-adjusting IoT systems.

What do you see as the biggest problems involving IoT deployments at large?

Many industry participants have put the platform “cart” before the use-case “horse.” That is, they have incorrectly focused on developing components of a solution and left it to the end customer to assemble all of the pieces into a usable system. This approach assumes a much more mature ecosystem than we currently have at this nascent stage of market development. As a result, many platform suppliers (note: here I use the term “platform” in a manner analogous to the PaaS definition, i.e., a structure upon which applications can be more readily developed) find themselves waiting for the real party to begin. While true IoT platforms are undoubtedly the direction in which the industry is headed (albeit with only 3-5 surviving players), it’s exceedingly difficult for a new technology to get started that way. More common is the development of highly-targeted but complete solutions that allow end customers to easily get started solving actual business problems.

What kind of policy changes or societal shifts do you think are needed for the Internet of Things?

If by “policy” we mean “government policy” then the answer is none. The best thing government can do to facilitate rapid innovation and growth in IoT is to stay out of the way, much as it did with the internet itself. Societal shifts will largely be the result of IoT, not something that is needed by IoT.

What is your advice to other industry professionals looking to deploy an IoT solution?

Most successful IoT implementations are not actually born as IoT initiatives. Instead, they are born out of efforts to improve asset uptime, reduce service and warranty costs, improve product design and reliability, increase machine or vehicle efficiency, etc. In other words, the genesis for all of these efforts is a significant business problem that needs to be solved. Frequently, as businesses analyze potential solutions, they begin to realize that device data can be extracted and analyzed. They discover patterns that humans were unable to detect and develop rules that can orchestrate actions in order to achieve the underlying business objective. In other words, they back into an IoT system. It is a mistake for industry professionals to deploy IoT solutions for the sake of IoT, despite the fact that many industry promoters recommend they do precisely that.