In 2020, a major consumer goods company approached KIS to improve the efficiency of their powdered soap manufacturing division. Their factory was already the most technologically advanced of its type in the world, and the company’s team was well-qualified to interpret and validate their data. But they were spending millions on materials and operating hours running test batches to formulate their products. The company knew there had to be a better approach. How could they accurately predict the quality of their product without wasting so much money? To find the answer, they turned to KIS.
consumer goods
3-6 months
machine learning, artificial intelligence, web services
Every soap product at the company has a unique density. Factors like temperature, water level, vacuum pressure, etc. all contribute to those densities.
To accurately predict how these variables affect each product’s density, we needed data from each production phase. Fortunately, the client had already outfitted their factory with sensors that took measurements at each phase and sent the data to a cloud.
Our challenge was to use this data to create three machine learning models that could reliably predict the product’s density after each phase. With these models, users could virtually create new products by adjusting input variables and validate existing product formulas. Without running physical test batches, the company would save time and money while simplifying the workflow.
To address the client’s problem, we broke the production process down into three distinct phases, each of which affects the product’s density differently:
Our plan was to create models to predict the product’s density after each phase.
At first glance, machine learning seemed like the perfect solution for the client’s problem. But we wanted to be sure.
Often, ML is like a hammer in search of a nail. Due to the buzz surrounding it, companies use ML to address problems otherwise easily solved by more conventional means. KIS isn’t one of those companies.
We began by analyzing the client’s data to make sure machine learning really was the best solution. The client sent KIS 18 months' worth of data on three soap products produced at their factory. Our team validated the data and looked for correlations that would help ML models accurately predict product density.
In the end, our ML experts found that the project was a perfect match for machine learning’s capabilities.
Now that we’d determined that machine learning was the way to go, we had to decide which ML models would work best for this application. There are hundreds of ML models available, and choosing the best one for a given project is no easy task.
Instead of trying different models randomly, our experts carefully deliberated which ML architectures and models would work best for the job and with the data provided by the factory sensors. After only a month spent validating the data and running benchmarks, we arrived at a model that predicted outcomes from the drying and mixing phases with a maximum accuracy of 92%! This far exceeded human capabilities and proved that our models could reliably replace physical testing.
But there was still one area where KIS saw room for improvement: the user experience.
As accurate as our new machine learning models were, they wouldn’t do the company much good if no one knew how to use them.
Our solution was an attractive, user-friendly interface that replicated the factory virtually. This digital twin of the factory allowed users to visualize the production processes, manipulate product formulas, and read the outcomes as easily as if they were running physical test batches.
By placing ourselves in the employees’ shoes, we came up with a solution that not only worked but made their jobs even easier.
.png)
With our models implemented, the factory’s efficiency increased. The company’s ability to validate their products was at an all-time high, and the client was no longer wasting valuable time, money, and materials by running test batches.
Want to know exactly how we did it? Keep reading to find out.
The digital twin UI tool uses an Angular application that receives all variable inputs. It then connects with a trained model binary through REST API call. Its design mirrors the physical factory in a virtual space, making it easy and intuitive for a non-technical user to interact with the ML model.
By taking a practical approach to this project, KIS implemented advanced machine learning technology to successfully solve the client’s production problem and simplify their factory’s workflow. Without jumping to flashy solutions or overcomplicating the issue, we addressed the problem with the exact level of complexity it demanded. Our team worked quickly and efficiently, creating ML models in under a month with enough accuracy to eliminate the need for physical testing at the factory.
As a result, these models saved the client from wasting materials and operation costs while providing employees with an easy-to-use UI that makes validating product formulas quick and easy.
The client was thrilled with the result, and they continue to work with us to this day. KIS is currently in the process of implementing the next model in the series which will no doubt continue to increase efficiency.
At KIS, we see ourselves as software generalists. We don’t go into any project with a single solution in mind. We take a holistic view of the problem and come up with the simplest solution to address it. In this case, machine learning was the perfect solution, and our ML experts were more than suited to the task. But, where a consultancy that only specializes in ML might have left the client with a mess of unwieldy tech, we had the expertise to implement a UI that made navigating the technology easy and intuitive. By maintaining a diverse team and keeping our tech solutions people-focused, we did everything in-house and solved the client's problem as quickly and sustainably as possible.
tech strategy & consulting.
We empower companies, invest in IT, provide tech training, create scalable squads, and deliver the right value.
data
management.
We handle data migration, conversion, analysis and integration, and database consolidation and management.
machine learning models.
We use AI algorithms and data to improve the accuracy of your analyses and help you make better predictions that guide decision-making processes.
web services development.
From discovery to delivery, we solve every issue, not just the basic ones.
