With no apologies for the word play, yet another digitalization opportunity is emerging to further aid our ability to simulate the real world virtually and help organizations and individuals make ‘better’ decisions.
What is a Digital Twin? A digital twin is a virtual representation, a model, of any physical asset, process, system or environment that is created to look and behave like its counterpart in the real world. These models use data inputs to mimic real world conditions and help design, development, operation, prediction and scenario testing.This arrangement is more cost effective than real world simulations, however the data modelling concept is not new as engineers have been using computational modeling since the 1960’s.
What’s Changed? The ability to leverage computational modelling has been accelerated by significant performance and cost improvements in data processing power, device and sensor technology. The move to constantly stream real-time data into a model makes the digital twin more dynamic.
Further twinning the digital twin with data and process mining – additionally boosted through the application of AI and machine learning – is another evolution that is gaining traction (PWC,”Twinning the digital twin with process mining: the right recipe for a truly connected supply chain”). This set-up can be an extremely valuable tool to help organizations generate insights on process gaps, bottlenecks and inefficiencies, and then simulate alternative scenarios.
The Crunch? Garbage in equals garbage out. Outside the cost of software, sensor /device hardware and IoT cloud capacity and connectivity, a significant amount of people time and effort is required to build and translate the digital twin into a meaningful model. The challenge for many organizations are the skills needed to comprehensively identify, structure, and map data in the context of the applicable process flow. Over simplification results in an inaccurate model, and over complexity typically confuses.
Forming a digital twin is therefore likely to more attractive for specific industries, for example, construction and manufacturing, where there is a more direct line of sight into the computational model from the start. For those organizations without a clear line of sight, the use case and digital twin ROI benefit may feel overwhelming. Accuracy is determined by a large quantity of good high quality data and where datasets contain critical errors, and/or miss key attributes; this can confuse baselines. This complication may discourage organizations from understanding the value of creating a digital twin.
AYSCO CONSULTING 