The Data Analytics Crossover. How Can Sports Data Analytics Concepts Like Moneyball and Marginal Gains Be Applied to the Process Plant Operation?

“People in both fields operate with beliefs and biases. To the extent you can eliminate both and replace them with data, you gain a clear advantage.”

― Michael Lewis, Moneyball: The Art of Winning an Unfair Game

This quotation wouldn’t be out of place in a morning meeting at a refinery.  How many times do we find teams arguing over whose spreadsheet has the right numbers and do you know any operators who change setpoints to their personal preferences during their shift? You certainly won’t find anyone’s 40 years of operational experience and tacit knowledge inside a data lake.    

Most sports fans and baseball fans in particular will know the story of Billy Beane in Moneyball, a book written by Michael Lewis. Beane was the GM at Oakland Athletics in baseball’s US Major League.  Beane challenged complacency and traditional methods for evaluating players by introducing a statistical, data-driven approach to team formation and selection. This approach revolutionized not just baseball, but the entire sports industry.

Letting someone else carry the shopping bags…

David Brailsford is the father of Marginal Gains theory and he implemented this philosophy when Head of the hugely successful British Cycling Team for track racing and Team Sky for road races.  The core principle of Brailsford’s marginal gains approach was that if you made lots of small improvements in many areas, you could roll up the small wins and generate a significant performance improvement. Brailsford looked at the traditional areas such as fitness and tactics but didn’t stop there, he looked at technological advances in micro components of the bicycles, helmet design for the best aerodynamics, and even race day bodysuits.

The obsession with the minutia even extended to pillow selection for the best night’s sleep and one of the riders even mentioned that he did not lift the suitcases into the car on the way to the airport to avoid the risk of a last-minute injury.

Everybody loves a good stat

There is an argument for comparing Marginal Gains to the concepts of Continuous Improvement and Kaizan, and a Moneyball approach facilitates the in-shift delivery of such gains and continuous improvement for the operational workforce.  The challenge for the industry lies in organizing the data landscape so that it can be easily accessed, mined, and analyzed by the plant operations teams.  To do this, they need an easy-to-learn and easy-to-use analytics platform that can connect to disparate plant data sources.

With plant operations, however, you must not put all of your baseballs in one “Data” basket, you do need to factor in and leverage the tacit knowledge of the engineers and operators. Digitizing this wisdom and knowledge can be achieved through electronic shift logs, data annotations, and journals.

By harnessing powerful self-service analytics tools such as Seeq and sports analytics concepts like Moneyball and Marginal Gains, the industry can benefit and continuously optimize its operations in a variety of areas, delivering incremental improvement over time and breakthrough improvements all at once.

Everyone can now knock it out of the park with Seeq

Seeq’s Workbench enables engineers to engage in comprehensive analytics, from data connection and cleansing to modeling and calculation, specifically tailored for time-series data in process manufacturing. This way, by tapping into plant data, companies can discover new ways to solve old problems and identify new problems they didn’t know they had. With Seeq’s self-service analytics platform, it’s all about getting those sought-after insights quickly and using those insights to make analytics-driven decisions that generate better business outcomes.

With Seeq, engineers are quickly transforming raw data into valuable information, and using Seeq to democratize plant data analytics means that every day, thousands of analytics-driven decisions are being made by the operational teams. These decisions are helping to increase yields, improve quality, reduce energy costs, prevent unplanned events and deviations, and resolve process issues.

This collection of use cases, happening throughout all shifts, means that continuous improvement is now being delivered in real time. It can be argued that the citizen approach and mass analytics being made by the workforce are essentially equivalent to Brailford’s concept of marginal gains, where the summation of many marginal gains produces a significant performance improvement.   

Home Run Use Cases

You don’t always have to sweat the small stuff, it should be noted that due to the relative immaturity of plant data analytics and data science in industry, there are often “Home Run” use cases that deliver extremely high value catches all at once.  The MOL Group, for example, recently presented a Seeq conference and they described how they were able to use Seeq to optimize the transition time between grades and minimize waste production across their plants, representing a 6 figure benefit per plant.  Parklands Burnaby refinery in Canada also presented how they used Seeq to reduce their flaring and associated emissions resulting in a $600,000 saving per year.

A 1% reduction in energy usage and maintenance costs has a significant impact on the balance sheet

Have you ever looked at your company’s financial quarterlies or annual report?   If you have then you will most likely see that the cost of energy and cost of maintenance are two of the highest operating expenditures for your business. Therefore, it is logical to start with marginal gains in these higher-cost areas of the business.

Where are the biggest lost opportunities?

Operating a process plant has different complexities compared to managing a sports team and sometimes it is not always obvious where gains can be made.   With sports, you either win, lose, or draw but with industrial plant operation, you may not even know when, where, or why you are losing, or perhaps your workforce does but there isn’t a culture of marginal gains or continuous improvement in place to address gaps in operation. Echoing the story of Moneyball, there is also a tradition of doing things the way they have always been done in the industry, and we often hear, “If it ain’t broke, why to fix it”.  Data and data analytics can help to overcome such complacency and change inertia and it normally only takes one successful use case to convert neigh-sayers into die-hard fans.  After completing a use case workshop and subsequent online test drive of Seeq, one of our customers recently exclaimed “Where have you been all my life!”…

The Team Huddle 

So data analytics can be applied to achieve gains for the industry, but how can the results of the analytics be communicated to the team?  Seeq Organizer steps up to the plate as the tool that ramps up collaboration among engineers, operators, and managers. Its role? To help assemble and disseminate analyses, reports, and dashboards. But it’s not just a feature; it’s a key player in ensuring that the valuable insights unearthed through data analytics don’t gather dust—they are actively shared and put to work across the organization.

In essence, Seeq Organizer facilitates cross-departmental communication and data collaboration, making sure everyone in the organization has access to the results of the analytics.  Seeq Organizer aims to turn the data analytics-driven results into a shared, rapidly available, and easy-to-consume asset.

Shifting Gears to special teams… Leveraging Data Science and Advanced Analytics with Seeq Data Lab

With the advent of data science and generative AI, the scope for applying Moneyball-type data analytics and pursuing marginal gains has increased and Seeq’s Data Lab is at the head of the peloton.   

Seeq Data Lab extends the workbench capabilities of Seeq into the realm of Data Science and Python libraries, offering process engineers and data scientists a platform for more advanced industrial analytics.  The data lab allows for more sophisticated, complex analytics and models, driving deeper insights and more refined improvements where routine analytics is required with a heavy computational load.

What does ChatGPT think about this?

[User:  Act like an industry consultant and describe how Brailsford’s Marginal Gains theory and Moneyball be applied to process industries using GenAI and Seeq

ChatGPT:  Combining Sir Dave Brailsford’s Marginal Gains theory, the Moneyball approach, GenAI, and Seeq in the context of process industries provides a comprehensive framework for continuous improvement, data-driven decision-making, and optimization. Here’s how you can integrate these elements:

Data Integration with Seeq:

• Moneyball Approach: Use Seeq to integrate data from various sources within your process industry, including sensors, equipment, and production systems.

• Marginal Gains: Apply Seeq to identify patterns, anomalies, and correlations in the data that may indicate areas for improvement. Continuously monitor and analyze data to uncover insights.

Process Optimization with GenAI:

• Moneyball Approach: Employ GenAI to analyze historical and real-time data, identifying critical variables that significantly impact process outcomes.

• Marginal Gains: Implement GenAI algorithms to optimize process parameters continuously. Fine-tune settings based on data insights to achieve incremental improvements in efficiency, quality, and resource utilization.

Predictive Analytics for Maintenance:

• Moneyball Approach: Use Seeq to monitor equipment performance and identify potential maintenance needs based on historical data.
• Marginal Gains: Implement predictive maintenance models with GenAI to anticipate equipment failures before they occur. This minimizes downtime and enhances the overall reliability of the production process.

Talent Optimization and Training:

• Moneyball Approach: Leverage Seeq to assess the performance of individuals and teams within your workforce.

• Marginal Gains: Utilize GenAI for personalized training programs based on individual and team performance data. Identify and nurture talent, optimizing the capabilities of the workforce over time.

Supply Chain Optimization:

• Moneyball Approach: Use Seeq to analyze historical supply chain data and identify areas for optimization.

• Marginal Gains: Apply GenAI to predict demand fluctuations and optimize inventory levels dynamically. Implement data-driven decisions to enhance supply chain efficiency and reduce costs.

Quality Control and Process Monitoring:

• Moneyball Approach: Employ Seeq for real-time monitoring of quality metrics and process performance.

• Marginal Gains: Use GenAI to continuously optimize quality control processes. Implement algorithms that adjust parameters based on real-time data to ensure consistent, high-quality outputs.

Continuous Improvement Culture:

• Moneyball Approach: Foster a culture of continuous improvement by encouraging teams to use Seeq to identify areas for enhancement.

• Marginal Gains: Utilize GenAI to track and measure the impact of small improvements over time. Encourage employees to contribute ideas for marginal gains, and leverage Seeq to visualize and communicate the cumulative effects.

Risk Mitigation and Decision Support:

• Moneyball Approach: Employ Seeq to assess historical data and identify potential risks in the process industry.

• Marginal Gains: Use GenAI for real-time risk prediction and mitigation. Implement data-driven decision support systems to respond proactively to emerging challenges.

Strategic Innovation:

• Moneyball Approach: Use Seeq to analyze industry trends and benchmark against competitors.

• Marginal Gains: Apply GenAI to identify opportunities for innovation within the process industry. Implement small, incremental changes that collectively contribute to a more competitive and efficient operation.

By integrating Seeq, GenAI, Brailsford’s Marginal Gains theory, and the Moneyball approach, process industries can create a holistic approach to continuous improvement, leveraging data insights for optimized operations, talent management, and strategic decision-making. Regularly assess the cumulative impact of small changes to drive long-term success.]

What do we say?

It is tempting to ask Chat GPT to write an entire article for you but we find it lacks detail and specific examples and it is yet to learn the art of incorporating puns and a play on words.  However, the results from the ChatGPT do align with our conclusions with one exception. 

ChatGPT seems to treat Seeq and GenAI as two different items but Seeq has recently issued a sneak peek into a GenAI extension where the features and benefits of GenAI can be exposed and utilized within the Seeq platform!