Sussex AI Seminar IX Harry McCarney: What we can learn from Digital Twins and Simulation - A pragmatic approach to
By: Aleks Kossowska
Last updated: Monday, 29 April 2024
Harry McCarney
Harry McCarney is MD of Hack&Craft ().
Title: What we can learn from Digital Twins and Simulation - A pragmatic approach to the epistemic status of insights gained from digital models.
Abstract:
Digital models, in which I include Simulations and Digital Twins, are increasingly essential to navigate complex global challenges such as supply chain planning and sustainability. My company, Hack and Craft, develops the Digital Twins and simulations critical to the global supply chain. In particular, our Digital Twin of the world’s largest pallet network is used to ensure the availability of approximately 80% of the world's consumer goods.
The ability to explore scenarios in digital replicas enables business and government leaders to de-risk initiatives by visualising potential impacts in a relatively low cost environment which does not disrupt real world operations. However, there is a lack of clarity and shared understanding about the nature of different kinds of digital models and the epistemic status of the insights they can offer. This uncertainty creates an unease about their use and ultimately holds back the adoption of these potentially transformative tools.
In this talk I will present a framework for classifying different types of digital models and the kinds of claims they make. This will require a pragmatic and accessible treatment of a range of issues in philosophy of science and epistemology, including differences between empirical and analytical claims, inductive versus deductive propositions, falsifiability, counterparts, the utility of thought experiments, and predictive versus explanatory models. The talk will bring these philosophical questions down to earth with real examples from Hack and Craft’s experience using simulation and digital twins to guide key decisions of global supply chain leaders.
Hopefully the talk will provide a starting point for further work which could rigorously situate the insights gained from digital models and provide a pragmatic framework for prudent and reflective use of these powerful tools.
Watch the recording of seminar given 17/04/2024: