Decision regression for modelling of supply chain resilience in interdependent networks: LNG case

As an application of advanced analytics (AA) in supply chains (SCs), to model supply chain resilience (SCR) of transactions, logistics, operations, etc., of such complex representation of networks, we propose a supervised machine learning approach as a predictive analytics decision regression modelling framework that uses a coefficient setup MIQP (mixed-integer quadratic programming) technique. This can determine optimisable surrogate models to correlate independent X variables (e.g., resistance and recovery of the SC resilience) to the dependent Y variable SCR considering a dynamic behaviour of the SC with lag- and dead-time. A novel methodology to quantify SCR, based on a tree of continuous x and binary y variables of resistance (avoidance and containment) and recovery (stabilisation and return), considers ad hoc relationships of x and y to be part of the SCR predictions in the machine learning MIQP identification method. Such SCR algebraic or analytical formulas obtained in this constrained decision regression approach (a type of predictive analytics) can be used in optimisation and control problems of prescriptive and detective analytics types of AA. The proposed model is applied in an oil and gas case pertaining to liquid natural gas (LNG) and its leaks known as boil-off gas (BOG). The vapors generated by the leaks or venting of BOG in the LNG supply chain reduces the SCR of this commodity. There are losses of materials, environmental impacts, reduction in the calorific value of the LNG to be re-gasified, potential safety issues, to name a few. This research aims to introduce a methodology to model and predict SCR in a general way (by a new design as example), but particularly, covers the digital transformation implementations that can potentially lead to an enhanced resilience in the supply chain of LNG towards the desired level of digital supply chain resilience (DSCR).