'Extra Sensor' Perception: How IoT Revolutionizes Service Management
Mark Twain once quipped that the difference between the perfect word to and the almost-perfect word is like the difference between lightning and the lightning bug. The general idea is that while language may be malleable, there's something to be said for the poignancy of specific terms.
The same could be said for sending the right technician to fix an industrial machine. Historically, service techs were Jacks of all trades. When a machine went down, you picked up the phone and called the agent, who typically sent out whomever was available at that moment.
Once on-site, the tech would need to do a detailed assessment of the machine in question. Which part appeared to be broken? What would need to be done to get it operational again? Ascertaining this information took time, research; possibly some calls to HQ for guidance.
The end result was decidedly hit-or-miss. Maybe you got lucky, and the technician really knew your particular machine. If the stars all aligned, he even had the right parts at-the-ready in his truck. A few twists of a screwdriver later, the machine was back online and running smoothly.
But if you were unlucky? Well, the tech shows up, spends an hour or more head-scratching, calling aimlessly to various parts manufacturers; waiting on hold. And once the problem is identified, guess what: the necessary parts are nowhere to be found today. Maybe next week?
In the manufacturing world, the impact of machinery downtime ranges from painful to disastrous. Depending upon the equipment in question, the dollars lost per hour can run from hundreds to tens of thousands or even more. Getting things fixed is a command performance.
Extra Sensors
Fast forward to the modern enterprise and we find a much different scenario. Thanks to the power of IoT technology (Internet of Things), companies can monitor any number of mechanical assets, practically any type of machine, from just about anywhere, even in real-time!
The technology is not new, but recent advances have dramatically improved the efficacy and functionality available. As IoT platforms mature, the relative cost of monitoring devices has gone down, while the ability to not only monitor, but manage devices remotely has grown remarkably.
Perhaps the most significant advance in the last couple years involves the use of machine learning and artificial intelligence (AI). These technologies can be used to dynamically and automatically slice and dice data to reveal meaningful insights, and trigger targeted actions.
One incredibly valuable use case for the confluence of AI and IoT involves predictive analytics. With enough training data over time, algorithms can actually predict when a piece of hardware will fail. This can happen via robust regression testing to find correlations between parts.
In such examples, in IoT solution can actually send a message to the business when a particular machine is about to fail! In these cases, the algorithms have found that, for example, when one part reaches a certain temperature, it's close to a failure point. Voila! Problem solved!
The Power of AI
But the most modern systems today go even further. Savvy organizations can leverage unstructured data about the service technicians themselves, to determine areas of expertise. This can be incredibly valuable when dealing with highly specialized, and expensive, machines.
For situations like these, an AI-enabled IoT solution can not only prevent the failure of a mission-critical machine, but also identify -- in real time -- who the optimal service technician is within range of the job. Take a moment to absorb how amazing that is!
The way it's done? With enough data about the technician, such as Human Resources data, employee profile information, but more importantly, transactional details gathered from production systems: the algorithms can identify expertise down to machine parts and job types.
An excellent source of information for this kind of algorithm would be the ticketing system used by a service provider. Such systems have both metadata and free text, both of which can be leveraged by AI to classify the experience of the tech with remarkable granularity.
And last but certainly not least? The parts! Far too many hours of possible production have been lost for want of the right parts! The ideal solution here is to have a failure act as the trigger to a purchase order that gets sent to the appropriate suppliers.
These types of dynamic frameworks can solve all the tricky, manual challenges that occur when a company needs to do maintenance or repairs on machinery. That allows the human resources to focus on the really fun, interesting and challenging work of bringing it all together.
While all of this might sound daunting, and certainly does require the appropriate technology, personnel and methods, just remember another of Twain's famous quotations: "The secret of getting ahead is getting started."