to Unlock Immediate Value with
In an increasingly competitive environment, driven by commoditizing product markets, evolving customer expectations, and shrinking margins, industrial machine builders are recognizing the significant contributions that the aftermarket business makes to overall company profits. To unlock more value from each service interaction, industrial machine builders are pursuing the technologies and tools that transform their service delivery—from reactive to proactive and preventative service models.
By leveraging connectivity via the industrial Internet of Things (IIoT), service organizations can harness data to predict failures before they occur. This capability transforms service outcomes, helping to increase uptime, accelerate time-to-resolution, reduce truck rolls, and improve the customer experience.
But some industrial machine builders may be reluctant or unsure about starting a predictive service program without in-house data science expertise, a large pool of historical data, and ample time to accumulate the necessary data on which to build models. So, how can your company expeditiously arrive at the next frontier of service?
3 KEys
Predictive Service
From years of design, testing, and field service procedures, your technicians have already accumulated deep expertise on the specifications of your machines. This knowledge, in combination with the real-time data you receive from your connected, deployed machines, can be utilized to make predictions.
Know when to anticipate a failure mode by building condition-based alarms based on comparisons between your real-time analytics from the field, existing engineering specifications, and historical performance data. Once your team is alerted that your specs have been violated––or will be––they can get ahead of customer downtime and bring in parts with long lead times
Predict Based on
What You Already Know
Applying What You Know: An Example
1
Is it true that the more data you collect from the beginning, the better your service? Not necessarily. Collecting data just for the sake of it just takes up storage in your database—and makes running your queries more complicated. But you can make smarter predictions based on what you learn from looking for patterns in the data you acquire over time.
Whether you have an established connectivity program or are newly connected, focus and refine your data collection efforts based on what is germane to a specific problem you want to address. Ask the right questions of your dataset to uncover answers, and with the help of artificial intelligence (AI) and machine learning, support your models becoming more accurate and effective over time.
Predict Based on
What You Learn
2
Today, simulation is more widely available and accessible than ever. Your engineers can create physics-based models to simulate all permutations of the real-world operating conditions and stresses that your machines can encounter in the field.
Not only does this help you identify and rectify design issues before a part ever goes into production, but using simulation, you can virtually operate the design under extreme conditions, in order to understand where predictive alerts are needed. You can even use data collected over time to make your simulations run better.
Predict Based on
What You Simulate
3
Your peers are leveraging connected machine data and the latest field service technologies to proactively prevent problems, avoid costly truck rolls, and accelerate time-to-resolution––and you can do it too. Now is the time to enable true predictive service to transform your costs and deliver superior customer outcomes.
Share this on social media
No matter where your company is in its connectivity journey, you can introduce predictive service now—and achieve a timely return on your IIoT investment.
Learn how to leverage your connected data and simultaneously apply these three key methods of prediction to your service strategy today.
Unlock Predictive Service
Predictive Service
Based on your design specifications, a deployed machine cannot perform effectively if its temperature goes out of range. From the incoming stream of field data, your technicians are alerted that the specs have exceeded the threshold for three consecutive readings. By comparing the connected data to existing specs, the technicians predict that a failure mode is pending and take action to prevent downtime.
Applying What You Learn: An Example
To determine when a particular part of your machine will fail, you must understand what the predictors of failure are—whether it’s the vibration levels, air pressure reading, or any other critical measurement. From their field service procedures, your technicians know the measurements have a significant bearing on a machine’s performance—and the collected data supports that it is a primary predictor of part failure. Collecting data on critical readings, combined with additional machine learning data, enables the continuous improvement of your model’s fidelity over time.
Applying What You Simulate: An Example
Instead of physically producing all the conditions and variables that your machines might encounter in the field, your engineers build a simulation model in the latest iteration of your design process. Not only does this simulation identify potential areas of failure before you even make a part, but it can inform the design of your warranty to help reduce costs.
21028
for Industrial Machine Builders
