The concept of the digital twin has become one of the defining ideas of modern manufacturing. By creating a virtual replica of a physical production system, engineers can monitor, analyse, and optimise operations without disrupting the real process. When combined with energy modelling, digital twins offer a powerful mechanism for identifying and eliminating waste. Swedish innovation agency Vinnova supported the Swedish contributions to the SPEAR project, which explored precisely this intersection of digital twin technology and energy optimisation.
Within the SPEAR project, the term "digital shadow" was used alongside "digital twin" to describe two distinct levels of virtual representation. A digital twin, in the SPEAR context, referred to a complete simulation model of a production system that could be used for offline planning, scenario testing, and virtual commissioning. A digital shadow, by contrast, was a continuously updating representation that ran in parallel with the actual plant, consuming real-time sensor data and mirroring the current state of the physical system.
From Static Models to Living Representations
Traditional energy audits produce a snapshot: a point-in-time assessment of how much energy a facility uses and where the biggest consumers are. While useful, these snapshots quickly become outdated as production schedules change, equipment degrades, or new products are introduced. The digital shadow approach developed in SPEAR addressed this limitation by creating a representation that evolved continuously alongside the real plant.
The implementation used embedded hardware-in-the-loop simulations running on cost-effective processing units. Real signals from the plant's programmable logic controllers fed directly into the simulation, keeping the digital shadow synchronised with actual operations. When a robot on the production line executed a welding sequence, the digital shadow simultaneously computed the predicted energy consumption of that sequence based on the component-level energy models. Any discrepancy between prediction and measurement triggered model refinement, gradually improving accuracy over time.
Automotive Production as Proving Ground
Volvo Car Corporation's body manufacturing facility served as one of the primary validation environments. The Swedish partners developed a complete digital twin of several production stations, incorporating robot movements, tool activations, conveyor operations, and associated energy flows. Using this twin, the consortium evaluated optimisation scenarios including adjusted robot trajectories, modified cycle timing, and alternative sequencing of energy-intensive operations.
The results were significant. Energy consumption at the tested stations dropped by 12%, and individual robots showed potential savings of up to 33% through optimised motion profiles. Importantly, these savings were achieved without any reduction in production throughput or quality. The digital twin allowed engineers to verify that proposed optimisations would not introduce unacceptable cycle time increases or create coordination conflicts between stations before any changes were made to the real line.
Scalability Across Industries
While the automotive use case demonstrated the approach most comprehensively, the SPEAR consortium also validated the methodology in forging, bakery production, and general manufacturing environments. In each case, the fundamental approach remained the same: build component-level energy models, compose them into a system-level digital twin, synchronise with real operations, and use the resulting data to drive optimisation algorithms.
The transferability of the approach across such different domains highlighted one of the project's core design principles. Rather than building sector-specific solutions, the consortium created a generic platform that could be adapted to any production environment where energy behaviour could be modelled at the component level. As manufacturing continues to face pressure to reduce both costs and environmental impact, this kind of flexible, evidence-based approach to energy management is likely to become increasingly standard practice.