Orchestrated intelligent supply chain optimizer
The present invention relates to systems and methods for intelligently optimizing supply chain is provided. In particular, the systems and methods provide the capability to configure supply chain systems so as to: balance between cost and service is optimised and profitability maximised; configure system parameters to respond to both current and future risks; ensure that variability is built into plans enabling maximised efficiency; and human error and bias are eliminated from the planning process such that pro-active rather than reactive behaviour becomes the norm.
1. An orchestrated Intelligent Supply Chain Optimizer comprising:
a Data Management Module as computer executable code, stored on non-transitory computer memory, which when executed by a computer system performs the step of configuring connections to data from a customer enterprise data system (EDS), wherein the customer data is received in at least one of an asynchronous process and a synchronous process, wherein the customer data includes fixed data related to an item, item parameter data and constraints, and wherein the customer data is incoming transformed through mathematical operations to make it usable by some parts of the computation system;
a Parameter and Model Archive database for storing the customer data;
a demand forecaster for consuming forecast and demand history time series data to generate a variation model of variations between sales forecasts and actual demand, generating bias and noise terms from the variation model, drawing a distribution of functions that have been constructed from past demand data, building a raw model based upon the distribution using a regression calculation with the bias and noise terms, and conditioning the raw model on forecast data to generate a demand forecast model;
a delivery performance modeler for consuming historical delivery performance data to generate a delivery model;
an Optimization Module as computer executable code, stored on non-transitory computer memory, which when executed by the computer system performs the step of segmenting products by defined business strategies and assigning different service level values based upon the segment, selecting a set of input feature vectors iteratively optimizing parameters for the set of input feature vectors using the demand forecast and delivery models, computing sensitivities for the set of input feature vectors by creating an internal model of behavior for each of the input feature vectors localized to the neighborhood of the optimized parameters, wherein the number of iterations is dependent upon computer processing capacity and desired accuracy for the optimization parameters; and
improving computational performance of the Optimization Module by having an adaptive approach to input feature vector selection by following contours of a fixed service level or utilizing a gradient based search algorithm to identify input feature vectors near local optima.
2. The Optimizer of claim 1 , wherein the Optimization Module comprises:
a Segmenter Adjudicator as computer executable code, stored on non-transitory computer memory, which when executed by the computer system performs the step of segmenting a plurality of products based on currently available data and for adjudicating updated segmentation with previous segmentation results;
a Strategic Constraints Definition Module as computer executable code, stored on non-transitory computer memory, which when executed by the computer system performs the step of identifying constraints on allowed supply chain configurations to ensure compliance with both physical limitations of systems and corporate governance and strategy;
a Supply Chain Attributes Definition Module as computer executable code, stored on non-transitory computer memory, which when executed by the computer system performs the step of characterizing the physical and performance characteristics of the relevant parts of a supply chain quantitatively; and
a Future Performance Predictor as computer executable code, stored on non-transitory computer memory, which when executed by the computer system performs the step of predicting future performance of the supply chain for each fixed set of operational parameters is predicted for the supply chain defined by the Supply Chain Attributes Definition Module.
3. The Optimizer of claim 2 , wherein the Future Performance Predictor is further configured to predict future performance of a supply chain network for different supply chain parameter settings based on a variety of supply chain network configuration assumptions, analysis of optimal supply chain parameter settings given both strategic objectives for individual products or groups of products and system-level constraints.
4. The Optimizer of claim 1 , wherein the Optimization Module is further configured to determine a current supply chain network using machine learning and AI models to recommend improvements upon an existing network and facilitate implementation of these improvements.
5. The Optimizer of claim 1 , wherein the optimization is performed on a node-by-node basis for the supply chain, and wherein the computation of a leaf node that is characterized by the delivery performance of a given upstream leaf node which defines the total transportation costs to the leaf node and manufacturing costs of a product, ignores intermediate nodes between the leaf node and the upstream node for the optimizing the supply chains.
6. A method for optimizing an Orchestrated Intelligent Supply Chain comprising:
retrieving customer data from an EDS asynchronously and/or synchronously and storing the customer data in a database;
selecting a set of input feature vectors;
consuming forecast and demand history time series data to generate a variation model of variations between sales forecasts and actual demand;
generating bias and noise terms from the variation model;
drawing a distribution of functions that have been constructed from past demand data;
building a raw model based upon the distribution using a regression calculation with the bias and noise terms;
conditioning the raw model on forecast data to generate a demand forecast model;
consuming historical delivery performance data to generate a delivery model;
iteratively applying the demand forecast and delivery models to customer data to generate supply chain optimization results by applying the demand forecast and delivery models to the selected feature vectors, wherein the optimization results include optimized parameters, wherein the number of iterations is dependent upon computer processing capacity and desired accuracy for the optimization results;
computing sensitivities for the set of input feature vectors by creating an internal model of behavior for each of the input feature vectors localized to the neighborhood of the optimized parameters;
generating display parameters for the optimization on an ongoing basis;
conditioning and presenting optimization results for optimal usefulness and impact, wherein the optimization results are remotely accessible by a user, and wherein prioritized results are transmitted to the user automatically; and
improving computational performance of a system for optimizing the Orchestrated Intelligent Supply Chain by having an adaptive approach to input feature vector selection by following contours of a fixed service level or utilizing a gradient based search algorithm to identify input feature vectors near local optima.
7. The method of claim 6 , wherein applying the AI models to the customer data of the Orchestrated Intelligent Supply Chain comprises:
segmenting a plurality of products based on currently available customer data;
adjudicating updated segmentation with previous segmentation results;
identifying constraints on allowed configurations of the Supply Chain to ensure compliance with both physical limitations of systems and corporate governance and strategy;
characterizing physical and performance characteristics of relevant parts of the supply chain network quantitatively; and
predicting future performance of the supply chain network for each fixed set of predicted operational parameters.
8. The method of claim 6 , wherein the presenting optimization results comprises:
generating a percentage dial visualization of the realized optimization;
generating a time series chart of the cost over time of the supply chain with an overlay of an optimized indicator; and
generating at least one heat map of two optimized parameters.
9. The method of claim 8 , wherein the at least one heat map includes reordering frequency on a first axis and safety stock levels on a second axis.
10. The method of claim 8 , wherein the at least one heat map includes a position of actual parameter settings.
11. The method of claim 8 , wherein the at least one heat map includes a position of a Naive theoretical calculation of a suboptimized solution for comparison against the two optimized parameters.
12. The method of claim 8 , wherein the at least one heat map includes at least one constraint overlay that limits the two parameter combinations allowed for optimization.
13. The method of claim 8 , wherein the at least one heat map includes a plurality of heat maps over multiple sections of the supply chain.
14. The method of claim 13 , further comprising receiving a manual selection of a parameter at a given node in the supply chain and automatically updating the plurality of heat maps located in downstream nodes.
15. The method of claim 6 , wherein the optimization is performed on a node-by-node basis for the supply chain, and wherein the computation of a leaf node that is characterized by the delivery performance of a given upstream leaf node which defines the total transportation costs to the leaf node and manufacturing costs of a product, ignores intermediate nodes between the leaf node and the upstream node for the optimizing the supply chains.