A leading manufacturer of paper-based cups, containers, and folding cartons, sought to improve their production process through advanced data analysis and visualization. With a mission to provide innovative sustainable packaging solutions and lower the amount of material waste during manufacturing processes, the client faced significant challenges in their data extraction, cleaning, and analysis. We based our cooperation on a need to optimize their production processes, reduce waste, and enhance overall operational efficiency.
Brief Summary
The client, a prominent manufacturer specializing in paper-based cups, containers, and folding cartons, embarked on a transformative journey to enhance their production processes using advanced data analysis and visualization, aiming to provide innovative sustainable packaging solutions and minimize material waste during manufacturing.
Their commitment to sustainability and waste management was reinforced through advanced data analysis and visualization, ensuring alignment with their operational and financial goals. The solution provided enabled the client to track waste effectively, reducing material costs and enhancing overall sustainability efforts.
Facing substantial challenges in data extraction, cleaning, and analysis, the client’s existing systems, including PostgreSQL, Syniti, and their AS400 systems, were insufficient in providing systematic and comprehensible data visualization, compounded by manual data entry errors which hindered accurate waste tracking.
In collaboration with the client, we meticulously redefined production data management, utilizing data cleaning techniques to correct anomalies and ensure data accuracy, thereby optimizing their production processes and enhancing operational efficiency.
NeuroForge provided the client with real-time, actionable insights, enabling proactive interventions and streamlined operations, resulting in significant waste reduction, improved data reliability, and more efficient decision-making.
Mission
The client has long been dedicated to reducing its environmental footprint, aligning its operational strategies with broader sustainability and waste management goals. By focusing on minimizing material waste during manufacturing, the client not only advances its commitment to sustainable practices but also ensures that its operations are economically efficient. Our collaboration exemplified how we can support these objectives by providing real-time insights into waste generation and enabling interventions that reduce waste and optimize resource utilization.
Through the implementation of waste data analysis and visualization, the client could seamlessly integrate sustainability into their core business processes. This contributed to meeting their sustainability targets and enhanced their operational and financial performance by reducing costs associated with material waste.
Challenges
The client’s production process involves multiple stages, where each step naturally generates some waste. They wanted to gain insights into their waste generation to further optimize it by lowering the amount. To achieve this, they built a first prototypical solution on top of PostgreSQL, Syniti, Refine, and PostgREST to analyze data from their AS400 systems. However, they faced several issues.
Firstly, the attempted solution had some data issues. The complexity of the processes within the AS400 led to a complex data schema. This meant that some parts of the system were hard to reason about. It was unclear if anomalies were due to data anomalies or due to programming mistakes.
Additionally, the application was not yet optimized for production usage; it existed as a prototype, but was not deployed to a productive system, which requires proper change and release management processes. The manual data entry process in the AS400 systems further complicated the functioning of the waste calculation, as workers could simply forget to input the necessary data into the system.
Despite these challenges, the client needed insights to fulfill their mission of responsible manufacturing by lowering the waste rates. They needed a solution that could clean existing data by implementing indicators to highlight potential (data entry and production) errors, process and store data effectively, and visualize data in user-friendly dashboards, all within the timeline set for this initiative.
Solution
Our collaboration with the client brought to life a solution designed to address the challenges they faced in data analysis and proper waste calculation. We conducted several interviews and workshops to understand the specific needs of their business and technical teams, ensuring the solution would be practical and aligned with their goals and work routine. To make that happen, we redefined the way production data was managed in PostgreSQL, enabling the client to gain deeper insights and optimize their production processes effectively.
The first step involved a meticulous assessment of the client’s existing systems and data flow. Their production data, primarily stored in AS400 systems and then copied to PostgreSQL for analysis, required significant preprocessing. As the data in AS400 was not intended for analysis, mainly due to the complexity of their processes, it was challenging to calculate accurate waste values.
Together with the client, our experts came up with simple yet powerful KPIs that could help identify real waste as well as data input errors. By highlighting these data issues, we empowered users to identify and address inaccuracies in their process documentation, rather than glossing over potential problems. For different types of products, we analyzed the processes and designed algorithms to calculate the amount of waste properly. With these algorithms, the system was able to calculate the input and output factors to arrive at a "waste" coefficient.
Redesigning the data models was crucial for improving data handling. We worked closely with the client’s technical team to map out each production step, identifying key input and output variables. The redesigned models incorporated these variables and established clear relationships between them. This process involved creating new data schemas in PostgreSQL, optimized for both storage efficiency and query performance. To increase performance even further while also making it easier to develop the production dashboard, we used cube.js to build a semantic data layer on top of our optimized PostgreSQL schema.
A central aspect of the solution was developing a custom dashboard tailored to the client’s specific needs. Utilizing modern front-end technologies like React.js and Refine, we created an intuitive interface that visualized real-time production data. The dashboard provided a comprehensive view of the client’s production processes, highlighting key performance indicators (KPIs) such as production output, waste generation, and efficiency metrics. Our solution enabled users to drill down into specific data points, allowing for detailed analysis and informed decision-making.
To transition from a development environment to a production-ready solution, we employed Docker for containerization and Docker Swarm for orchestration. This approach ensured the deployment process was streamlined and automated, reducing the risk of human error and improving system reliability. Continuous Integration/Continuous Deployment (CI/CD) pipelines were established using Azure DevOps pipelines, enabling rapid testing and deployment of updates. This automation expedited the development process and ensured the system remained robust and adaptable to changing requirements.
A critical component of the solution was ensuring the client’s staff could effectively utilize the new system. We provided basic training sessions, demonstrating key ways to use the dashboard and offering simple explanations in concise documents. Additionally, we established ongoing support and maintenance services, ensuring the client could rely on NeuroForge for any future technical assistance.
This collaborative effort not only met the client’s immediate needs, but also provided a scalable framework for future growth and innovation. The real-time dashboard provided actionable insights, enabling the client’s decision makers to identify and address inefficiencies promptly. The redesigned data models and automated data processing pipeline ensured the client’s production data was accurate and readily available for analysis.
Outcome
The partnership with NeuroForge changed the client’s approach to production data analysis. The new system provided them with a live, real-time dashboard that significantly improved their ability to monitor and optimize their production processes. The results included:
Enhanced ability to track and reduce waste at each production stage, leading to cost savings.
Automated data cleansing and anomaly detection improved the reliability of production data.
Streamlined processes and real-time insights led to more efficient operations and better decision-making.
The success of this project not only met but exceeded the client’s expectations. We believe that this collaboration underscores our commitment and expertise in developing custom AI and data solutions that deliver tangible business benefits.
If you are looking to enhance your production processes with advanced data analytics, NeuroForge has the expertise to help you achieve your goals. Contact us to discover how we can empower your business with innovative, custom-tailored solutions.