Developing IoT solutions for existing product lines is increasingly important to industrial companies of all sizes and scopes, primarily because they’re seen as important value-adds without the need to reinvent the wheel itself.
One example: manufacturers are adding sensors to their devices in order to provide customers with a preventative maintenance service – the product is monitored and, when necessary, serviced, without inconveniencing the customer. The result: no disruptive field failures, the product’s service life is extended, and by extension the brand enjoys a stronger affinity from the customer.
That’s the theory, anyway.
IoT Potential vs. Reality
Because we live in an era with virtually unlimited potential for imagining digital tools and the connective networks that bind them, industry is faced with a difficult decision: while IoT devices sound good in theory, how do you know if an IoT investment actually makes sense?
To answer that question, we recommend answering four key questions about the users (and uses) of the IoT product prior to actual development and deployment. We’ve bundled these into what we call the 4W Framework (Who, When, Why, and What). As in:
- Who is the consumer of the data that flows from the device?
- When do they take an action?
- Why do they take action?
- What do they do?
Now let’s look at an actual 4W scenario played out in the case of a predictive maintenance solution.
| WHO | The operator of the equipment |
| WHEN | As soon as they receive a ‘flag’ showing an imminent failure |
| WHY | It’s a part of their job responsibility |
| WHAT | They start an internal maintenance process with a request |
Immediately we can start to ask some key questions about the proposed IoT deployment:
- Are we sure that we know ‘Who’ is correctly identified? Is it the operator of the machine or some other person that receives the alert?
- Do we understand the time between the imminent failure notification and actual failure?
- Have we assessed the actual job responsibility and ensured that they are responsible for taking action?
- Do we understand the customer’s internal maintenance process?
Expanding the Picture
All of the above is related to the real time data flowing from the IoT device. But to understand the full deployment we need to look further out in time and other applications of the data.
Let’s look at the 4W for a collection of data from multiple devices deployed with a customer, perhaps over a period of months.
| WHO | Maintenance leader |
| WHEN | Monthly |
| WHY | Responsibility for machine uptime / OEE performance improvements |
| WHAT | Aggregate data to show patterns in failures, response time of the maintenance process, improvements in field failure rates |
So, for the same deployment but this time across a collection of IoT devices, a different organizational role uses the device data for different reasons.
It’s worth noting here that in the case of preventive maintenance the IoT deployment success is intimately tied to the customer’s maintenance process and is almost certainly linked to core metrics about Overall Equipment Effectiveness (OEE). This is likely to be the core of the value proposition around a preventive maintenance IoT solution.
Simply stated, if the response time of the customer maintenance process is longer than the time before failure predicted by the IoT solution then there will be no benefit to the customer – they will still suffer from field failures.
Finally, let’s look at ‘4W’ for the longer-term collection of data from the IoT devices deployed for preventative maintenance.
| WHO | Supplier data analysts and product engineers |
| WHEN | Multi-month, multi-year |
| WHY | Assessing performance of the predictive model |
| WHAT | Adjust the algorithm predicting failure, input to the next generation product road map |
Once again the people who use the data flowing from the devices are a different group and they are looking at the data for different reasons.
Summing Up
Now let’s consider the deployment of an IoT solution for predictive maintenance across all stated roles and purposes:
| Real Time | Medium Term | Long Term | |
|---|---|---|---|
| WHO | Operator of the equipment | Maintenance leader | Supplier data analysts and product engineers |
| WHEN | As soon as they receive a ‘flag’ showing an imminent failure | Monthly | Multi-month, multi-year |
| WHY | It’s a part of their job responsibility | Responsibility for machine uptime / OEE performance improvements | Assessing performance of the predictive model |
| WHAT | They start an internal maintenance process with a request | Aggregate data to show patterns in failures, response time of the maintenance process, improvements in field failure rates | Adjust the algorithm predicting failure, input to the next generation product road map |
In developing an IoT solution for predictive maintenance it is important to recognise that:
- There are different users of the data that come from the devices
- The value that they derive from the data is different
- The ways in which the data is presented to them will differ
- The value of the data that flows from the device is intimately linked to the performance of an internal customer business process
- The IoT value proposition can only be enhanced by having access to the large, long-term data set
- There are multiple potential customers for the IoT solution who have different needs
The bottom line, is that the answers to the 4W Framework are an important part of the solution design process that will uncover the true users (around whose needs the solution needs to be designed); the customer’s processes (whose performance may affect the delivery of value from the IoT solution); how the true value of the solution can be presented to customers; and the ways in which the data needs to be collected and presented.
