Coca-Cola İçecek Improves Operational Performance Using AWS IoT SiteWise

The purpose of the digital twin was to automate the shop floor and provide CCI with a holistic view of its manufacturing processes. With this information, CCI could improve line and asset utilization operations, and its operators could increase measurement accuracy and deploy preventative maintenance when necessary. “If we can locate failures and other maintenance issues before they happen, we can keep the plant up and running at all times and improve our utilization,” says Suheyla Er Aksoy, asset optimization digital technology leader at CCI. A complete digital plant replica in the cloud would also help the company gain value through advanced analytics, artificial intelligence, and near-real-time asset monitoring.

To power the digital twin, CCI needed to connect data points from its assets and industrial IoT devices to the applications it needed to develop. Searching for a solution, the company turned to AWS. “We knew this was going to be a comprehensive solution that involved multiple aspects, such as IoT data modeling, processing, and performance,” says Elif Ege, digital twin product manager at CCI. “AWS offered a flexible suite of services and provided the level of support we needed to build a scalable and configurable solution.”

CCI chose AWS Professional Services—a global team of experts that helps businesses realize their desired outcomes on AWS—to build a scalable analytics solution using AWS IoT SiteWise, a managed service that makes it simple to collect, store, organize, and consume data from industrial equipment at scale to augment decision-making with data. As part of CCI’s commitment to diversity, a diverse team spearheaded the project. “Diversity is critical to our sustainability, and inclusion is at the epicenter of our culture,” says Aksoy. “We were able to form a highly competent and diverse leadership team to drive this product.”

During the project’s first phase, CCI used AWS Professional Services to build a solution for its clean-in-place (CIP) process, a critical sanitation process in the food and beverage industry that cleans interior surfaces of production lines and equipment without disassembly. “Our CIP is an everyday production process and crucial for industry hygiene and quality requirements,” says Burcu Hacioglu, product owner for the digital twin product team at CCI. “We wanted to improve process efficiency by avoiding errors in time measurement, which could lead to CCI plants overusing energy and water.”

CCI needed a way to collect and process enormous amounts of industrial data as well as build digital models of CIP assets and processes. To ingest equipment data for processing, CCI used AWS IoT SiteWise, which runs on AWS IoT Greengrass, an IoT open-source edge runtime and cloud service that helps developers build, deploy, and manage device software. AWS IoT SiteWise ingests a large amount of industrial data from CCI plants and enables operators to monitor processes at the edge using Grafana dashboards, an open-source analytics and interactive visualization web application. Amazon DynamoDB, a fast and flexible NoSQL database service for any scale, is used for state machine processing and calculates and compares operational data.

Using this solution, CCI operators receive access to digital representations of company assets and gain visibility into the CIP process in near real time. The IoT data is further enriched in the cloud using business rules and process-related information, which is powered by AWS services such as AWS IoT Analytics, a fully managed service that operationalizes analyses and scales automatically to support up to petabytes of IoT data. This enriched data is then stored in an industrial data lake and served to CCI operators and company stakeholders through Amazon Athena, an interactive query service that makes it easy to analyze such data on AWS using standard SQL.

CCI built the CIP digital solution in 2 months. “Working efficiently as a lean, agile team alongside AWS Professional Services helped accelerate our time to market,” Ege says. Within 4 months of deployment, CCI identified over 30 improvement opportunities that resulted in annual savings of 20 percent on electricity and 9 percent on water. Additionally, CCI estimates it will save 34 days of processing time per year using the digital twin solution, which will reduce overhead costs and further increase efficiency. “Between 2021 and 2023, we expect to regain 200 production days by implementing the digital twin of the CIP process,” says Aksoy. “We also expect to optimize our filler-mixer machines, which have a high impact on our line utilization.”

By the end of February 2021, CCI deployed the CIP solution to four systems and expanded the digital twin to include filler-mixer machines in three of its plants. Using these solutions will help identify machine failures and improve production-line performance. To achieve its goals for asset optimization, human-machine collaboration, improved line utilization, and sustainability, CCI is working toward automating manufacturing processes in all 26 plants.

In the future, CCI will continue to transform its bottling operations at scale through near-real-time monitoring, process analytics, and artificial intelligence implementation. As CCI extends its digital twin across plants and lines, it will continue to develop machine learning models on AWS to drive intelligence on the shop floor. “With the digital twin, we engineered a data- and technology-based asset optimization approach,” says Aksoy. “This accelerates our efforts toward autonomous smart manufacturing augmenting our teams with IoT data and AI.”