Artificial intelligence (AI) and additive manufacturing (AM) are poised to transform industries, but their combined potential for sustainable practices, particularly regarding material waste and circularity, remains largely untapped. This research explores the intersection of AI, specifically machine learning (ML), and AM, focusing on powder bed AM technology, through the lens of material waste reduction and circularity.

AM's inherent digital nature and vast data generation create a fertile ground for ML algorithms. The research delves into the types and sources of data – process parameters, material properties, build performance – and potential variabilities in experimental and simulation data. It critically assesses the applicability of this data to inform ML models for material waste optimization.

Beyond existing applications like defect detection and process optimization, the paper proposes innovative ideas of a ML-driven closed-loop feedback systems: Adaptively adjusting process parameters based on real-time data to minimise material waste and ensure part quality. Predicting and preventing machine failures to reduce wasted material from aborted builds.

Utilizing ML to predict material behaviour and optimize printing parameters for recycled or scrap metal powder reducing reliance on virgin resources.

The research envisions a future where AI fully realises its potential in AM, driving significant progress towards a circular economy in manufacturing. This future encompasses closed-loop material streams, seamless integration of recycled materials into AM processes, guided by AI-powered decision-making.

The data-driven optimization can personalise printing experiences based on individual part requirements, minimizing material waste and maximising resource efficiency. AI-driven design for lightweight structures and functional optimization, reducing overall material consumption.

This research highlights the immense potential of AI in AM for promoting material sustainability and circularity in manufacturing. By harnessing the power of data and intelligent algorithms, AM can have more responsible resource-efficient future.