The ceramic manufacturing proces are energy-intensive and prone to unexpected equipment failures that cause costly downtime. KAIZEN leverages cutting-edge AI to help companies in ceramic industry operate more efficiently, sustainably and reliably.
Case Study
Optimizing energy use in Kale Ceramics manufacturing
Our work process
At KAIZEN, we follow a systematic process to deliver optimized AI solutions for our manufacturing clients. Our work takes place across four key phases:
Evaluation
We identify pain points through interviews and a full data inventory and maturity assessment.
Strategy
We determine the quickest ways to deliver value, set up teams, budget, and define success metrics.
Development
Our experts build custom analytics solutions from the ground up.
Deployment
We ensure smooth transition to production, monitoring performance and business impact.
The Problem
Ceramics manufacturing is an energy-heavy industry, with natural gas use makes up for the majority of expenses. Making efficient use of this resource is critical, not only for the companies bottom line, but also to reduce the carbon footprint of the industry.
- Natural gas expenses are one of the largest costs in ceramic tile manufacturing.
- Kale's gas furnaces and other equipment were consuming 15% excess energy due to inefficiencies.
- This significantly increased costs and carbon footprint. Difficult to calculate excess use without real-time monitoring.
The Solution
Data: We worked with 12 months of sensor data from multiple production lines in a single factory. Sensor data consisted of temperature, pressure and energy consumption values from individual pieces of equipment, as well as data related to controllable set parameters of the system. Insight: Analysis revealed that for certain time periods, large discrepencies can exist between desired parameters and observed values. These anomalies were calculated to cause excess energy usage of up to 15%.
Modeling: To remedy this excess, modeling needs to produce two outputs. First, it needs to identify anomalies in energy usage. Then, it needs to identify the proper set parameters to bring energy usage within desired limits while staying within operation constraints.
- Installed AI-based systems to monitor all equipment and detect anomalies.
- Built optimization models to recommend optimal equipment settings to minimize gas usage.
- Created energy analytics dashboards providing real-time visibility into usage.
Deployement
Our models were deployed in the cloud according to the needs of the customer. Integration with the customer’s existing data infrastructure was performed through an automated process.
We built an online, real-time dashboard, which displays critical system state and inform ground personnel of anomalies when they happen, recommending optimal set values for energy use.
The work is currently under live testing, and is expected to reduce energy expenses by 5% overall. This translates to 1,100 tons of yearly CO2 emissions for a single factory, a significant gain for an industry which is under great pressure to reduce their carbon footprint.
- Reduced natural gas consumption by 10%, saving millions in expenses annually.
- Lowered CO2 emissions by 1,100 tons per factory per year.
- Optimized performance without impacting production output.
- On track to meet sustainability targets years ahead of schedule.
Why KAIZEN?
Our Team
