In my years covering companies that try to add intelligence in the form of sensors and analytics to their existing operations, I often see those companies go down a similar path. They start with a pilot project and a sense of what they want to measure, but once they get started, realize that they might be measuring the wrong thing.
The path is difficult, requiring technical expertise, data science, and business acumen. And for most businesses, it’s increasingly paved with consultants.
For example, as part of an optimization project, a company might speed up the pace of a pick-and-place machine by focusing on the picking algorithm, only to realize that the manufacturing line slows down when it reaches a spot welder because the two pieces require more welding time. And after adjusting the welding time, the fix might uncover new issues later on.
The process of improving manufacturing and other industrial operations is similar to a science experiment. You start with a hypothesis, measure your results, and adjust accordingly. When done well, it can lead to new business metrics that more accurately help a company improve its returns or perhaps solve another business problem. When done poorly, it can lead to a game of whack-a-mole as companies and consultants try to figure out what the data is telling them.
Vinay Nathan, CEO of industrial IoT platform company Altizon, describes these challenges as “the floating bottleneck.” The idea is that companies start with a theory about a machine that might be causing an issue or slowing down a process, instrument that machine with sensors, and then use analytics to improve the process.
Sometimes that’s all it takes. But more often than not, the bottleneck simply shows up elsewhere in the plant’s operation, leading to more instrumentation, another round of data and analysis, and more consulting hours. And because every IoT deployment is unique, figuring out a problem in one industry or on one class of devices doesn’t mean a company or consultant will be able to apply that knowledge to another one.
And so even tech-focused IoT firms have consulting divisions. What surprised me in my conversation with Nathan was that his company — which is a traditional IoT platform provider that offers connectivity, software, and integrations between different machines, software programs, and dashboards — has a stable of consultants that augment the work done by other consulting firms hired by the client.
A typical Altizon customer hires the company along with a systems integrator or consulting firm, such as Accenture or Deloitte. The management consultants then work with the Altizon consultants to merge the business goals with the technical challenges. This may seem like way too many consultants, but thanks to its complexity, the Industrial IoT is increasingly a rich source of revenue for professional services.
Because the problems are fairly new and unique to each customer, and because the technology infrastructure isn’t yet set in stone, professional services hold most of these implementations together. Even traditional manufacturing companies are building professional services teams to help their own plants, as well as to help sell their equipment or services to others. Both Emerson and Honeywell have launched such divisions in recent years.
In addition to the business and technology consultants, there are often data science consultants who work in the same field as the customer and who also help to set metrics and build the models needed to identify the hypothesized bottlenecks.
More than ever, it seems the industrial IoT is consultants all the way down.