Blumberg wrote, “IoT generates an extensive amount of real-time data, some of which is unstructured. In order to make use of this data in any meaningful way, a service provider will need to employ “data scientists.” This is true, to some extent. What we hear from most manufacturers is that they are drowning in analytical data; one mentioned that they receive approximately 3GB of data daily, from each machine in the field, yet they struggle to analyze even 1% of it. Many companies are collecting reams of data without a clear understanding of what they are looking for.
The Challenge With Collecting IoT Data
Does this mean that service providers need to employ data scientists, as Blumberg predicts? Maybe. The belief is that, with proper analysis, important patterns will emerge from the data that indicate equipment health and maintenance requirements. But that requires additional technology that can sift through the unstructured data and experienced personnel who can interpret the results. On the other hand, manufacturers and service providers can tap into a large pool of equipment experts, their call center and service personnel, to gather the kind of information that is directly relevant to improved performance and uptime. By bringing together fault isolation technology and data downloads, much of the clutter and noise of IoT can be removed revealing the root cause of equipment problems, even without “data scientists.”
Blumberg also predicts that there will be a need for more field service engineers, even though “IoT has the ability to improve the percentage of service events that are resolved remotely without dispatching a FSE.” He explains: “First, FSEs will be required to deploy IoT solutions. Second, FSEs will be needed to provide onsite diagnostics and troubleshooting when remote resolutions prove ineffective.”
There is some room for healthy debate about these statements. Can IoT actually improve remote resolution? We have heard from our customers and prospects that more data doesn’t necessarily help product support perform more effectively. In many cases, IoT has been seen to increase the number of “false positives,” meaning the on-board diagnostics signalled a problem when nothing was wrong with the equipment, which results in additional service calls. More FSEs may be needed, at least initially, because even when remote support works companies still send technicians on-site about 65% of the time to diagnose and repair equipment.
How OEMs and Operators Can Better Understand Data
Blumberg does not mention the impact that IoT has on product engineering. When engineering gets accurate feedback from the field it helps them design more reliable equipment. Unfortunately, IoT systems collect a variety of data that needs to be analyzed and interpreted, as compared to guided troubleshooting systems, which generate rich dashboards of failure modes, frequencies, and trends. With such dashboards on hand, OEMs can quickly recognize and respond to emerging issues before they escalate into customer support disasters. The connection between them is that guided diagnostic systems would treat signals detected in IoT data as symptoms, and not as conclusions. Pairing guided troubleshooting with IoT may prove to be the “killer app” that helps OEMs and operators to quickly discern data sequences that indicate a potential product or component failure; furthermore, when troubleshooting and diagnostic data is collected from a fleet of machines it can reveal emerging failure trends that require escalation and corrective action.
Manufacturers are trying to determine if the PLM/SLM-fueled promise of better reliability through IoT is real. (I.e., equipment quality will improve if you just add more sensors, collect more data, and run more analysis). Without using field experience to sanity check all that data, customers, service technicians and engineers can’t make better decisions about how to operate, repair and design equipment.