Growing analytics maturity to accelerate value
- Posted by Edwin van Dijk
You might have made your first steps into the world of data analytics to access the promise of Big Data, Industry 4.0 and the Industrial Internet of Things. You now know the technology is available and you’re enthusiastic about data analytics – but you’re still missing something that will make this work deliver the value you are seeking. If this sounds familiar, then understanding the current analytics maturity of your organization could be the key to determining the next steps that will bring you success.
Big data broken down
Process manufacturing companies are faced with enormous amounts of data from all kind of sources and sensors. This data may be structured or unstructured, depending on the source (such as customer information, email correspondence, lab reports, time series process data, device usage, etc.), and require different ways of turning data into actionable information. Of all this data, the most crucial for the profitability of the factory is the time-series based process data captured in your historian. Focusing on this process data is a good place to start your industrial analytics journey, as Marc Pijpers of Sitech pointed out in our previous webinar.
In 2010, Thomas Davenport wrote the book “Analytics at Work” in which he described his analytics maturity model. This model shows 5 maturity levels and for each level Davenport indicated the characteristics for the people in the organization, the processes and the technology used to benefit from the data available.
Based on the characteristics of these levels, Bert Baeck asked the audience of his webinar “Enabling Digital Transformation” to indicate the maturity level of their organization. The results were surprising: 78% of the audience indicated that they would classify their organization as Analytical Laggard up to Analytical Amateur. Only 12% was beyond that level.
It is certain that many organizations already have some executive awareness for data analytics – but it is also clear that their workforces are still not experiencing the benefits of being fully empowered by analytics. Often the reason for this is that while analytics are in place, they are not accessible to all stakeholders, but are instead restricted to the domain of analytics experts.
Democratizing data analytics
Currently there are two major approaches for data analytics: model based and self-service analytics. For model based analytics, data scientists are typically required. These analytics specialists use work bench tools to build an analytics model to answer key questions. This model is fed data from various sources, and the results are interpreted by the data scientists. While effective, this approach does not extend the daily benefits of analytics insights to the entire workforce.
Self-service analytics are designed to help subject matter experts, such as process engineers, control room employees and operations field engineers to use analytics themselves. Self-service analytics allow this entire group of people to directly dive into the time-series data and interpret asset and process behavior, leveraging their own expertise. This sharing of analytics insights is known as “the democratization of data analytics”, and it is one of the most far-reaching benefits of new technologies.
With modern self-service analytics tools, the employees with the most operational knowledge can:
- find root causes for process anomalies quickly within hours,
- monitor the process and asset performance and even
- predict performance and send early warnings if deviation of the golden fingerprint is expected.
“We need to empower the subject matter experts of the industry with analytics. With TrendMiner, process engineers, technologists, run engineers and operators can unlock the power of the industrial data for their day-to-day questions.”
– Fabrice Leclercq, Rotating Machinery Engineer TOTAL Refining & Chemicals
How to determine your next steps
No matter the analytics maturity level of your organization today, it is possible to improve. All that is needed to get started is a good understanding of the current situation and a clear idea of your objectives.
Assessing your current situation can be done based on “gut feeling” or with a more scientific approach, such as using an analytics maturity assessment tool. Once this initial assessment has been done, the next actions for deployment, managing change and enabling various groups to start working with analytics can be defined within your journey to digital transformation. As soon the workforce has adopted the analytics solution, they can start contributing to the organization’s overall profitability at all levels. The key is to get started in a way that delivers benefits from the start.
Interested in learning more? Watch the video
Learn how to make your analytics implementation a success. Hear key considerations for analytics projects and how to avoid barriers to success, plus how to chart an effective roadmap to uncover value in your existing data.
- Edwin van Dijk