MENU
Regardless of which industry you’re in, competing effectively often involves much more complexity than one would imagine.
In today’s market, organizations who want to use data to their advantage must know how to efficiently sift through all levels of detail in their data to provide leaders with answers to critical business questions.
But having the right data is only part of the equation. Much goes on behind the scenes before you can provide answers with speed and accuracy to enable more effective decision making.
This case study highlights how a leading U.S. stainless-steel manufacturer used technology and data to model one of their most critical business processes.
For steel makers, being able to accurately model and manage input costs for raw materials (also called unprocessed materials) can make a substantial impact on the overall profitability of their business.
This manufacturer has a digitalization team in charge of expanding the use of manufacturing and production data to make better operational decisions and improve financial performance of the company.
They specifically needed assistance building a modeling tool to help model and analyze costs of different unprocessed material inputs in the stainless process.
They would use this information to better manage their suppliers, the use of different input materials, and the overall production of different grades of steel, among other factors.
The group selected Onebridge to be their partner, and gave us the following objectives:
The manufacturer’s specific challenge, though different from other industries, is very relatable. To understand why this organization needed a raw-material modeling tool, it’s helpful to know that there are many grades of stainless steel, each with its own “recipe” of required metal percentages that are combined to create a final product.
The metallurgical characteristics, along with material costs, are the main factors that must be considered when looking at the overall cost of material used to make the final product.
Scrap metal (which has widely varied metallurgical content) is a major expense for stainless steel manufacturers. As the amount of each element in the scrap metal changes, it must be balanced with elements from other sources to achieve the desired target content for the end product.
The calculations to achieve the correct percentages are complex, potentially lengthy, and must be accurate within a fraction of a percent. Our client needed us to help design a user interface that could break down the complexity in a way that different users with various depths of technical knowledge could easily understand what they are seeing.
Our initial goal was to deliver a minimum viable modeling-and-balancing algorithm in a short period of time to help our client validate and quantify possible cost savings that could be realized by using different amounts and types of scrap to meet their production needs.
This algorithm would be part of an unprocessed-materials optimization tool we’d develop to provide our client’s stakeholders with a single source of information regarding actual raw material usage and cost over time and modeled scenarios using different assumptions.
The Onebridge team building the solution was uniquely qualified to meet our client’s needs. Though we were hired for application development and UX design, our work necessitated that we develop deep expertise in steel making and the key drivers that led to success in the process.
We already brought to the table vast experience in Agile project delivery and development, along with custom UI design and development. Additionally, team members possessed specific expertise in material science, algorithm analysis and extensions, and material optimization and balancing logic.
Working with the client’s digitalization team, we selected the tools and processes that would best align with their skillsets. We decided we’d use their existing tools. That would be the most sensible way to enable their team members to maintain and improve the solution as needed after we delivered the initial iteration and version of the tool.
The client had already partially completed some of the necessary calculations in an algorithm using a combination of stored procedures, SQL code, and other scripting approaches. Their existing reporting was based on complex Excel spreadsheets.
We used an Agile software delivery framework to collaborate with their team to refine business requirements, prioritize key functionality and capabilities for the final product, and test and validate the usefulness of those capabilities.
Throughout the 20-week initiative, we conducted daily standups (that included both Onebridge and client stakeholders) to review the status of stories managed in the Jira planning tool.
We worked within the existing SQL Server environment and utilized the capabilities of Python to create logic and a balancing algorithm to model different scenarios of unprocessed-material inputs that could be used to produce the desired end-product that was needed.
We iteratively added new business rules, balancing approaches, and re-processing steps to make the chemistry balancing algorithm more robust and complete.
In addition, we analyzed the cost changes (increases and reductions) across 24 scenarios after ensuring the raw materials were successfully balanced within the target element-content parameters of the final production grades.
Our team then incorporated the use of Dash into front-end dashboards and the user interface. They were created to report and visualize critical information relating to the material planning and usage process. The SQL code, combined with the rewritten user interface in a server-based application, enabled a wider range of analysts and engineers to make use of the results.
The final output, the unprocessed-materials optimization tool, consisted of a front-end, web-accessible interface that allows users to visualize, model, and assess different scenarios to guide decisions on material usage and supplier management.
The new dashboards are bulleted below. Each offers the ability to drill down to get further information.
The true value of the dashboards is that they provided the solution to our client’s challenge, which was the need to get timely, accurate answers and insights into critical questions like:
Our project empowered our client to be able to determine the opportunities to adjust raw-material content percentages in the raw-material scrap used as an input in the production process. We helped their digitalization team:
As a result, we increased the company’s agility and accuracy in decision making. That, in turn, yielded additional benefits like cost reduction and greater efficiency.
Our client appreciated the dedicated development environment, calling it very professional, and said they were thrilled with the balancing algorithm results, remarking the dashboards look fantastic.
The real proof of success was that the manufacturer was able to identify potential savings opportunities of approximately 0.5% on the raw materials cost based on different input chemistries. This equates to a possible monthly savings of $800,000 on raw material purchases.
Additionally, as users worked with the unprocessed-materials optimization tool, they gained a common understanding about raw-material purchasing and steel-production. That newfound, shared understanding helped align the organization to tackle upcoming projects and priorities.
Finally, throughout the initiative, we bolstered their team’s confidence in their own software development, and they were left with hands-on knowledge of Agile development done right.
“You and the team have been a pleasure to work with. Thank you for your patience as we have shifted direction countless times over the course of this project, as well as your dedication and persistence in seeing it through. We will continue to use the framework that has been established to make impactful business decisions.”
- Client Team feedback
