How to create value with just one metric!

The operating hours metric is simple but also extremely relevant for OEMs, dealers, and machine owners. Our example represents machines that are working in the construction industry, operating worldwide with thousands of machines. These machines can be organized by model.

This monthly average operation hours are important and are aggregated from the raw operational hours. To draw the most meaningful conclusions, the data should be collected and viewed over an extended period of time. The longer the observation period, the more reliable it is for providing insights.

In Figure 2, the x-axis shows the number of active days for each machine, and the y-axis shows the average machine operating hours per day. Note that each point represents a single machine. The point at the bottom left shows a machine that is active only one day for one hour, whereas the machine on the top right was active every day for 30 days, averaging 22 hours per day. 

Another important thing to note is that data points can overlap. An analysis of data over time will provide a better representation of the number of machines in a given plot. In Figure 2, the aggregated data is organized into a visually meaningful format by creating different clusters that are color-coded. Clustering is important because it helps organize similar behaving objects together. Here, we see three distinct clusters of machines, the green one, the orange one, and the purple one.

Figure 2 also shows which machines are working more than eight to 10 hours a day with a pink line. Those machines above the line are assumed to be working as part of a shift operation and exposed to considerably more use. To further analyze the working hours of the machines, three additional categories are shown in red, green, and blue.

OEMs and dealers can gain multiple insights from a single metric, depending on their goals and point of view based on their role in the business, as salesperson or service manager, for example. In the case of the metric regarding machine use, the answers to questions like these can reveal benefits for your business in terms of efficiency, profitability, and customer satisfaction.

Perhaps the data shows the same type of machine being used in different ways and at different times. This can mean that some machines are being over-utilized while others are under-utilized. Machines that are used more often will age faster and require more maintenance and repair. This example is more likely in the construction industry, where it's not so easy to ensure that machines will be used in a balanced way. For example, they'll be moved around from site to site or set aside for a while until the next phase of the job is ready. The ability to see which machines are working harder than others creates the possibility of intervening to balance workloads with better planning or by substituting a different machine that can do more in the same amount of time.

Figure 2 shows the use of machines over a period of time. Sometimes, a low-end machine that can do everything is constantly in use, cannibalizing the market of higher-end machines. The overuse of low-end machines can be an indicator of this situation and can give OEMs the insight necessary to make adjustments in the next model’s design. Similarly, when it comes to vehicles that are used every day, OEMs can design accordingly. For example, the seats should be as ergonomic as possible and the cab should have space for additional equipment. The models in the upper right of the char will require exterior lightning as well because they are used for shift work, around the clock. In contrast, the machines that only move once a week will require fewer features. With data that reveals how often machines are used or how they are used, it is possible to optimize for specific applications and so models don’t get in each other's way.

The analysis of machine data in Figures 4 and Figure 5 can have a positive impact on sales. Imagine a salesperson, looking at data for a certain group of customers. If he can see that a certain machine model is used a lot more often than the OEM intended with its design or if a machine owner is using a high-end machine for simple tasks, he could convince the customer to buy a model with a different configuration—one that’s better for the work it's performing. Using data to enable personalized buying recommendations is a very powerful way to build customer loyalty and understanding how machines will be used is the key to a successful recommendation.

Data about usage can affect the kind of after-sales service OEMs and dealers decide to provide their customers. From the data in Figure 2, we know that the average machine works five days a week, about seven hours a day. But that’s only the average machine. What about the others?  

It doesn't make sense to offer the same service package for every machine, regardless of their use.  

Data can also help with the prioritization of service requests and the cost of service contracts. For example, the hardworking machines in the upper right of the graph need to be serviced more often. Maybe it makes sense to offer weekly or bi-weekly service for these types of machines, whereas the machines in the lower-left corner need service much less often. The service contracts for these machines can be priced accordingly with the OEM or dealer being compensated fairly for the work involved or the machine owner not being overcharged, depending on the machine.

Usage data as shown in our examples can help OEMs and dealers predict which parts will need replacement and which machines are likely to need maintenance before others. This is not unlike what various automobile manufacturers are doing to bring their customers in for maintenance based on real-life vehicle usage data. The ability to reliably predict which machines will require maintenance and replacement parts can help OEMs and dealers achieve greater efficiency by anticipating their needs for staffing and having parts in stock, and planning accordingly.