• Home
• /
• Opinion

AI must join forces with materials science to save the planet

Nature is hardly able to regenerate, being weighed down by the burdensome waste. We therefore need to act now.

The recent heat wave has scorched the lives of North Americans. This surge in temperatures in unlikely places is being linked with climate change. It has killed an estimated 1 billion marine lives on the coast of Vancouver. Through centuries of human activity-led waste generation, we have tested the limits of nature’s ability to self-correct. Nature is hardly able to regenerate, being weighed down by the burdensome waste. We therefore need to act now.

The impetus to move to a circular economy is gaining momentum across the world. Circular economy strives to design out waste from products. Apple, which happens to be the third largest manufacturing company in the world, revealed that merely 10 per cent of materials in Apple products shipped in 2019 was from recycled or renewable sources. This puts the spotlight on the materials we use to build products.

Today, materials science has a fundamental role to play in saving the planet as we face resource depletion and environment degradation.

Materials have been so elementary to the evolution of the human civilisation that entire eras were named after them such as the Bronze Age and the Iron Age. Materials science is as old as alchemy.

Materials science owes its humble origins to alchemy once alchemists started separating materials by using a distillation apparatus. The science evolved as an interdisciplinary offspring of Chemistry and Physics.

Fast forward to 2021, materials science is faced with a new challenge as some of its fundamental principles are being questioned.

Materials that make up our everyday products need to be reinvented so that they can be recovered and reused and stay away from landfills.

We are learning that waste is a design flaw right from the point when science is applied to materials. For example, the main hurdle in reusing and recycling metal is the alloying elements and additives used to provide materials with the requisite properties. This makes recycling of mixed material streams more difficult. There are ample number of facilities that segregate timber and steel from comingled waste. But the presence of copper and tin in steel makes it difficult to recycle steel. One of the suggestions is to examine the microstructure of the metal to better understand its grain and texture that are linked to the specific properties of the metal. We need to think of recycling material at its design stage.

Companies are using carbon data for product design and development decisions. The largest manufacturing organisation, Volkswagen Group has identified 16 high risk materials including battery raw materials such as cobalt, graphite, nickel, and lithium, which it aims to recycle in the future.

Apple has identified 14 materials that have the highest impact on its products and the environment, which include rare earth elements, aluminium, lithium and so forth to be prioritised for a 100 per cent recycled and renewable supply chain. Like Apple, most organisation are willy-nilly reexamining their product materials and looking for alternatives. It is not easy to wade through the vast amounts of data for actionable deductions.

For this, materials science needs to team up with artificial intelligence. In 2011, former US president Barack Obama had commissioned a nationwide project called Materials Genome Initiative. The initiative used AI to map millions of all possible combinations of chemical elements. This resulted in an enormous database allowing scientists to mix and match elements to create new ones.

Furthering this initiative, a team of researchers from US universities such as Cornell are using AI along with first principles quantum physics, experimental materials synthesis, processing, and characterisation to arrive at new materials knowledge. They are using a multi-agent system called the Scientific Autonomous Reasoning Agent (SARA) to find new materials and their synthesis parameters.

We are on the cusp of monumental shifts as businesses reexamine their decades of unintended contribution to global waste and climate change. If we are to save the planet, two technologies — one ancient and the other emergent — need to join forces. Deep Neural Networks need to go above and beyond, to not just discover materials but also reassess the consumption of rare materials within product life cycles. AI has tremendous capacity to accelerate the discovery of materials that support Circular Economy, and we are just getting started.

Shalini Verma is CEO of PIVOT Technologies