Digital Twin And AI Framework Helps Cut Waste And Energy Use, Study Finds
A reviewed study finds that digital twins and AI helped reduce waste by 27% and energy consumption by 32%, offering a scalable model for multiple industries.
A peer-reviewed study has demonstrated how digital twin technology combined with artificial intelligence (AI) and IoT sensors can help organisations reduce waste, improve energy efficiency, and unlock new circular economy opportunities.
The research, led by San Antonio-based engineer and researcher Shubham Gupta, was published in the 2025 edition of MDPI’s Proceedings and presented at the inaugural SUSTENS Meeting in June 2025. Although the framework was tested in an electronics manufacturing facility, the findings suggest it can be adapted to a wide range of industries, including healthcare, construction, agriculture, and food processing.
The study is among the first to provide independently verified, multi-dimensional performance data from a live industrial implementation of a digital twin circular economy framework, rather than a controlled simulation or theoretical model.
Pilot Project Delivered Broad Operational Improvements
The year-long implementation generated measurable improvements across multiple performance indicators.
According to the study, material waste fell by 27%, while energy consumption per unit dropped by 32%. Carbon emissions declined by 27%, maintenance downtime was reduced by more than half, and scrap rates decreased by 57%. The facility also increased recycled material content by 34% without affecting product quality.
In addition, the system identified 12 industrial symbiosis opportunities, enabling waste materials to be redirected as inputs for other internal processes or external partners. The framework also achieved 98% material traceability across the full product lifecycle, a level of visibility that allowed loss points to be identified and quantified with a precision that manual monitoring systems cannot match.
AI-Powered Digital Twin Moves Beyond Monitoring
Most industrial operations manage data in disconnected systems, making it difficult to understand how decisions in one area affect performance in another.
"Most facilities operate with fragmented data," Gupta says.
The framework addresses this by creating a real-time virtual replica of the production environment, integrating material flows, energy use, equipment performance, and quality metrics into a single model.
On top of this digital twin, machine learning models continuously analyse operational data to predict defects, assess recycled material quality, and optimise production schedules.
"The AI layer in this framework is a prediction and optimisation engine. It tells you what is about to happen and what you should do about it before it does."
This predictive capability allows managers to act before waste is created rather than responding after inefficiencies have already occurred.
Circular Economy Shifts from Cost Centre to Value Driver
One of the most commercially significant findings was the identification of 12 industrial symbiosis opportunities.
Industrial symbiosis means finding a use for one facility's waste stream as another facility's input."
By analysing waste stream composition, timing, and quality in real time, the framework highlighted practical opportunities to reduce disposal costs and generate new revenue streams.
The result is a different perspective on sustainability, where circularity becomes a source of operational value rather than a compliance requirement.
Applications Extend Well Beyond Manufacturing
The research has already gained recognition outside the manufacturing sector. A 2026 review on digital twins for sustainable membrane technologies, authored by researchers from institutions including the University of Edinburgh, Aston University, and the Italian National Research Council, cited the study. The inclusion signals that the framework’s architecture, which applies AI-driven optimisation to complex material and energy flows, is being recognised as a foundational methodology rather than a sector-specific application. A critical review is a curated synthesis of what defines the current state of a field; being cited places the work in the literature that shapes how subsequent researchers approach the problem.
The broader relevance lies in the fact that many industries face similar challenges: complex resource flows, hidden inefficiencies, and disconnected data.
"The framework transfers wherever three conditions exist: physical processes that generate data, resource flows that can be measured, and decisions that can be improved by better prediction."
Healthcare, construction, agriculture, and food systems are among the sectors identified as having strong potential for adoption.
As cloud-based AI infrastructure becomes more accessible, the study suggests that the key challenge is no longer technical feasibility, but organisational readiness to use data-driven tools to improve efficiency and sustainability.
(This copy has been produced by the Infotainment Desk)
Before You Go
Breaking: Muslim Side to Challenge MP High Court Verdict in Supreme Court Soon
Top Headlines
