4 minute read

Role of Al on the road to sustainable aluminium production

Role of AI on the road to sustainable aluminium production

By Quantillion Technologies*

Industry 4.0 is currently one of the most talked about topics in the aluminium industry, with such elements as automation, robotics, IIoT, ML, AI and AR leading the way towards the ‘smelters of the future’. Next to that Environmental, Social, and Governance (ESG) practices are becoming increasingly important, with the push towards sustainability coming from customers, governments, and potential investors. Getting hold of clean energy is only one side of this journey. The other side concerns process optimization opportunities arising from new technical developments.

At Quantillion, we strongly believe that making use of Industry 4.0 technologies is not only complementary to, but necessary for achieving sustainable production. Without an Industry 4.0 approach, it will simply not be possible to deliver these long-term goals with existing potline technology and control practices alone.

At Quantillion we have been using AI and a digital twin of smelter operations to optimise production processes. We see production environments as a cyberphysical area where humans, machines and equipment need to work together in the most optimal way. These optimisation gains are directly related to the increases in sustainability of operations. But how do you do it? Where do you start? How can we leverage the power of AI on the road to sustainable aluminium production?

In this article we will showcase an example of an AI solution which helped in improving sustainability of industrial operations as well as explain the steps that need to be taken when it comes to implementation of AI-based solutions in the heavy industry.

Hilbrand Kuiken, CEO Quantillion Technologies

At Quantillion, we are putting fast and powerful decision-making algorithms at the heart of production processes. Our products make it easy for operators and machines to make smart decisions, fast.

Hilbrand is the CEO of Quantillion. He is an entrepreneur focusing on data intelligence that enables real time autonomous decision making in complex environments. Hilbrand loves working with passionate people in creating real value for the clients. He believes data is not about analytics but about doing!

FAST. SMART. SUSTAINABLE.

ARTIFICIAL INTELLIGENCE

Artifi cial intelligence (AI), also known as machine intelligence, is an advanced technology that has already started to fundamentally transform many industries, including the heavy industry.

Based on a study published in scientifi c journal Nature Communications in 2020, AI could help achieve as much as 79% of Sustainable Development Goals. It has already proven itself in several other industries and is now becoming an important contibutor to solving a number of issues that are critical for sustainable metal production.

SUSTAINABILITY GOALS

The heavy industry has sustainability at the top of the agenda!

Net zero

A transformation that affects all countries and their industrial sectors (including aluminium), as per the 2050 Paris Agreement.

Decarbonisation

Aluminium customers are setting bold decarbonisation targets with a long-term goal of creating a CO2-free production process.

Circularity

Recapturing materials and well sorted scrap can improve the circularity of aluminium.

SHIFTING FROM A ZERO SUM GAME TO A NON-ZERO SUM GAME

Since sustainability goals are not individual goals it is crucial to change the way we work towards them and more importantly the way we think about these goals. � We have to address the sustainability challenges from total value chain perspective. � We should look for optimisation benefits through collaborating on a cluster level across value chain and industries. � We should not think of AI as the ultimate solution, but as one of the several means to achieve the goal.

EXAMPLE: AI HELPS TO REDUCE HEAT LOSSES

For an industrial cluster of companies in the Netherlands Quantillion developed a decision making model to reduce the total heat loss. By collaborating over multiple nodes, we were able to reduce the total emission from heat loss by 6%. It shows how non-optimal individual actions can be improved by making use of reinforcement learning. � The AI model was 20 times faster than a liner one making it more suitable for the real operational use. � Collaborative decision making outperformed individual decision making. � No infrastructural changes were needed. � The benefit was shared amongst multiple players and not only on the individual level.

WHAT OTHER APPLICATIONS ARE THERE FOR AI-BASED SOLUTIONS IN FACILITATING SUSTAINABLE METAL PRODUCTION?

WHAT ARE THE STEPS FOR SUCCESSFUL AI-BASED SOLUTION IMPLEMENTATION?

Output prediction and demand forecasting to eliminate excessive material use and/or waste.

Identification of opportunities for emission reduction.

Precise, real-time action recommendations to achieve a balance in energy use. Early error detection/ prediction in the casting process to reduce metal scrap. *If you would like to know what Quantillion’s AI solutions can mean for your plant and how to get there reach out to us!

info@quantillion.io

+31 85 060 52 88

quantillion.io

Quantillion originally started with automating transport logistics within primary aluminium. Since then we have grown to be a strong integration provider acting as a key link between the smelters and their numerous equipment suppliers.

The journey that we go through with the smelters consists of the following main steps:

1. Simulation of the current environment and future processes. 2. Preparation for IT integration of multiple types of equipment and manufacturing execution system. 3. Roll out and deployment of most suitable (AI) solution(s). 4. Further connection of other processes and equipment.