Patent Application
20080172262
1. Field of the Invention
The present invention generally relates to disaster prediction and mitigation planning and to disaster impact assessment.
2. Background Description
The frequency of natural and manmade disasters appears to be increasing globally. Examples of natural disasters include hurricanes, earthquakes, and pandemics. Examples of manmade disasters include drastic changes in economic conditions and geopolitical tensions leading to widespread labor unrest and war.
The impact of disasters on businesses grows as businesses become more globally integrated and interdependent, increasing businesses' reliance on partners and economies around the world. There has also been a trend towards reducing or eliminating redundancy in business systems and processes as companies have strived to reduce operating costs. This has in many cases left businesses more vulnerable to the risk of disaster-related disruption of their activities.
Traditionally, technologies used to predict the physical dynamics of disasters have been developed and utilized separately from technologies used to conduct and assess business operations. As a result, firms have not had access to fully adequate tools for integrating disaster planning into businesses processes.
The present invention bridges the divide between disaster prediction and business planning by facilitating the translation of physical and other effects from a disaster into dollars-and-cents impact on a business. In this way, a description of the impact of a disaster on the world at large can be reduced to measurable operational and financial implications for a specific enterprise. The present invention also allows a user to evaluate the costs and benefits of various disaster mitigation plans and/or policies and to understand the combined effects of multiple mitigation plans. This is achieved through the systematic analysis of multiple disaster scenarios.
This invention can be used to assist business leaders in assessing the business impact of a potential disaster. The main objective of the model is to quantify the impact of a potential disaster to the business, including the effect of government and/or business mitigation actions. The model accomplishes this by analyzing the potential impact of a disaster on factors such as business operations, physical and information technology (IT) infrastructure, company employees, customers, suppliers, business partners, revenues, costs and customer service levels. Importantly, these analyses can be used to understand and demonstrate the impact that the disaster has on a company as it evolves over one or more time periods, and over one or more geographical locations. Therefore, this invention is able to capture dependencies/correlations that may exist over one or more time periods and across one or more geographical locations, where geographical dependencies may exist within a time period and/or across time periods. It also evaluates the impact of various mitigation plans on factors such as business operations, physical and information technology (IT) infrastructure, company employees, customers, suppliers, business partners, revenues, costs and customer service levels over one or more time periods and one or more geographical locations.
The present invention comprises one or more of the following: a disaster dynamics calculator, an infrastructure factors calculator, an economic factors calculator, a behavioral factors calculator and a business performance calculator.
The present invention thus provides a computer-implemented method, a system, and a machine-readable medium for instructing a computer to estimate the business impact and risk associated with a disaster by:
Simulation may be used (i) to compute the global dynamics of a disaster and/or (ii) to measure the psychological, economic and physical impact of a disaster on a firm and/or the effects of various mitigation plans. A set of parameters and/or actions may be employed to characterize the mitigation plans. Examples of mitigation plan actions include keeping a safety stock of inventory, distribution of vaccines, cross training of employees, negotiating disaster clauses in supplier/customer contracts, and closing of a company site. Examples of mitigation plan parameters include the level of extra inventory to stock, the effectiveness of an evacuation or vaccination strategy, the starting and ending periods for the implementation of the plan, and the location(s) in which the plan is to be implemented.
The optimization of mitigation plans may involve (i) modifications to the structure of the dependencies between two or more of the firm's suppliers, business partners, customers, physical and IT infrastructure, and employees and/or (ii) modifications to the detailed parameters of the plans (iii) modifications to the durations of the plans, including the starting and ending period of the plans. The nature of these modifications may be determined through the design of one or more experiments which systematically explores the space of feasible parameters and actions, and identifies the combination of parameters and actions that optimizes the firm's performance with respect to one or more business objectives. Mitigation plans that are not controlled by the firm (i.e., government mitigation plans) may be imposed as constraints in the model.
The operational impact of a disaster on a firm may be measured in terms of available resources (e.g., people, materials, physical or IT infrastructure) allocated to demanded business processes, products and services. The detailed relationships and dependencies between available resources and demanded business processes, products and services may be captured by detailed enterprise dependency networks spanning multiple geographies and lines of business and/or global logistic networks.
The allocation of available resources to demanded business processes, products and services may be optimized using mathematical models, algorithms, and/or simulation. The demand for business processes, products, and/or services in each economic sector and line of business may be forecast using mathematical models that are sensitive to the effects of a disaster on various sectors of the economy and the availability of demand-generating and demand-sustaining resources (e.g., sales people) in each line of business.
A method or system for estimating the business impact of a disaster on a firm according to the present invention may therefore comprise:
In addition, or alternatively, a method or system for estimating the business impact of a disaster on a firm according to the present invention may comprise:
It is recognized that machine-readable instructions may be stored on a machine-readable medium to instruct a computer or other data processing apparatus to perform steps according to the method of the present invention.
The foregoing and other objects, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, in which:
Referring now to the drawings, and more particularly to
The disaster dynamics calculator 100 computes one or more scenarios of how one or more disasters will evolve over time. This is achieved by utilizing models (e.g., systems dynamics, logic, regression) that capture the ‘physics’ of the disaster. For example, in the case of a pandemic, the disaster calculator computes the change in the number of susceptible, exposed, infected and recovered people in multiple geographical locations over a time horizon (e.g., 1 year). The computed values may be provided for multiple time periods within this horizon.
The output from the disaster dynamics calculator 100 may be used as input to an infrastructure factors calculator 102, an economics factors calculator 106 and a behavioral factors calculator 104. The infrastructure factors calculator 102 computes the predicted effect that the disaster dynamics may have on the availability of infrastructural elements such as buildings, electricity, water, internet connectivity and ground transportation networks at different geographical locations.
The economics factors calculator 106 computes the predicted effect that the disaster dynamics may have on key economic indicators (e.g., gross domestic product, demand for services/products by industry sector) for different geographical locations.
The behavioral factors calculator 104 computes the predicted social and psychological effects (e.g., fear of becoming infected during a pandemic, staying home from work to care for family members, social distancing, rioting, looting, political unrest, lowered morale) that the disaster dynamics may have on people in different geographical locations. The predictions computed by the infrastructure, economic and behavioral factors calculators 102, 104, 106 may be provided for a specific time horizon, and for various time steps within this horizon.
Additionally, there may be dependencies between the infrastructure, economic and behavioral factors. For example, if people decide to avoid going to work, then this may affect infrastructure that require humans to maintain/operate them. As another example, if the economy experiences a downturn and the unemployment index rises, people may decide not to purchase certain goods/services, which may further hurt the economy. Therefore, the output from one of these calculators may be used as input to another. Eventually, some equilibrium state may be achieved or feedback tolerances/stopping rules may be implemented in the invention to prevent the occurrence of infinite feedback loops. The precise sequencing of the calculations and the feedback of data between these calculators can be controlled by a simulation manager.
The results from the direct infrastructure, economic and behavioral factors calculators 102, 104, 106 may be fed back into the disaster dynamics calculator 100 to affect the evolution of the disaster. For example, in the case of pandemic, as the infected population grows, fear may cause people to stay home from work and distance themselves from others. Such behavioral factors may impact the dynamics of the pandemic in future periods by slowing the spread of the disease. Eventually, some equilibrium state may be achieved or feedback tolerances/stopping rules may be implemented in the invention to prevent the occurrence of infinite feedback loops. The precise sequencing of the calculations and the feedback of data between these calculators can be controlled by a simulation manager.
The final results from the infrastructure, economic and behavioral factors calculators 102, 104, 106 are fed to a business performance calculator 108, which computes the effect of the disaster on business performance measures. Examples of such measures include revenue, profit, cost and service level, over the given time horizon. Therefore, the business performance calculator 108 is able to determine the impact various infrastructure, economic and behavioral factors on specific business/enterprise operations.
Each of the disaster dynamics, infrastructure factors, economic factors, behavioral factors and business performance calculators 102, 104, 106, including any sub-calculators that they may comprise, may be calibrated by the user through a set of input parameters 110 that the user can control.
Mitigation plans 110 may be input to any one of the calculators 102, 104, 106. These mitigation plans 110 may be instigated by a either individuals and/or groups, including governments, businesses, and/or international organizations. Mitigation plans 110 may have the effect of modifying (typically improving) the disaster dynamics (e.g., distribution of vaccines may reduce the infection rate in the event of a pandemic). Mitigation plans 110 may also have the effect of modifying the impact of the disaster on behavioral factors (e.g., government announcements reassuring the public in the event of a terrorist threat may help to boost public morale). Mitigation plans 110 may also have the effect of modifying the impact of the disaster on economic factors (e.g., a government can implement farming subsidies or low interest loans in the event of a drought or flood to stabilize the agricultural economy). Finally, mitigation plans 110 may also have the effect of modifying the effects that a disaster has on business performance (e.g., purchasing of insurance will mitigate financial losses if infrastructure is destroyed, cross training workers may improve customer service if there is a general labor shortage, and securing of alternative suppliers of raw materials will mitigate supply shortages).
Mitigation plans 110 may be implemented by various parties (e.g., a business, a government, etc.). It is possible that the effects of mitigation plans that are put forth by different parties, or even by the same party, may not always be in alignment. For example, in the event of a pandemic, a business may offer financial incentives to encourage healthy workers to show up at work. This plan would mitigate the impact of the disaster on worker absenteeism in the short term but may increase the risk of these workers becoming infected in the long term. At the same time, a government or local authority may provide education to the public encouraging them to stay at home. This plan may mitigate the spread of the disease but would increase worker absenteeism in the short term. This invention can be used to determine trade-offs between competing mitigation plans 110, as well as to determine the combined effects of complementary mitigation plans (i.e., plans whose effects are in alignment).
The resource availability calculator 206 may receive input 202 from the infrastructure, economic and behavioral factors calculators and produces a prediction of the availability of the human, material and infrastructural (e,g., information technology, electricity, facilities) resources that the business will have access to over the same planning horizon. These resources may be characterized by the geographical locations from which they are obtained, as well as by their physical or other properties (e.g., price and perishability in the case of material resource; wages and skills in the case of human resources). The resource availability predicted by the calculator may be a modification of a baseline resource availability that is provided as an input parameter to the resource availability calculator 206. This baseline represents the availability of resources under non-disaster conditions.
The demand forecasting calculator 204 may receive data 202 from the infrastructure, economic and behavioral factors calculators as well as the resource availability calculator 206 and produces a demand forecast for the products and services that the business offers to its customers/clients over a given planning horizon. This demand forecast may, for example, be expressed in terms of one or more of the following: demand volume (which may be stated in dollars, product units or full time equivalent employees) by day/week/month/quarter, customer type, customer location and industry sector. The demand forecast may depend on resource availability. For example, a reduced sales force may result in a lower demand forecast. The demand forecast produced may be a modification of a baseline demand forecast that is provided as an input parameter to the demand forecasting calculator 204, as shown in
The business resource dependency network 210 contains captures the relationships between each human, material and infrastructural resource and each product and/or service demanded. These relationships may be described at various levels of granularity, and relationships may depend on the location of resource, location of customers, customer account number, among other things. These relationships may also be tiered in the sense that a product/service may depend on the availability of one or more resources, which may in turn depend on the availability of one or more resources, and so on.
The resource allocation calculator 208 takes as input the business resource dependency network 210, the demand forecast and resource availability prediction. It determines an allocation of available resources to demanded products and/or services, taking into account resource requirement constraints, as defined by the dependency network. The allocation algorithm used by the resource allocation calculator 208 may be designed to prioritize certain products or services over other products or services. It may also prioritize the allocation of certain resources or other resources. In general, it may also optimize some performance measure such as maximizing revenue or minimizing cost.
The financial impact calculator 212 takes the results of the resource allocation calculator 208 as input and determines the expected cash flow, revenue, profit, cost and other financial indicators over the planning horizon. This calculator may take into account late payments, payment defaults, billing cycles, labor costs, contract payment structures, investment portfolios, costs of mitigation plans, exchange rates, interest rates and taxation, among other things.
After mitigation plans and all input parameters have been provided, the invention calculates the dynamics of the disaster in step 305. Subsequently, the invention calculates the impact of the disaster dynamics on infrastructural factors in step 306. If the infrastructure factors can influence the evolution of the disaster, as determined in step 307, then the invention re-computes the disaster dynamics, step 305, based on the updated infrastructure factor values. The iteration between the infrastructure calculator, as shown in step 306, and the disaster dynamics calculator, as shown in step 305, may continue until one or more user defined tolerance parameters are met. When the tolerance parameter(s) is (are) met, step 307, then the invention proceeds to calculate the behavioral effects of the disaster, as shown in step 308. Similarly, iteration between the behavioral effects calculator, as shown in step 308, and the disaster dynamics calculator, as shown in step 305, may continue until one or more user defined tolerance parameters, which may be different from the aforementioned tolerance parameters, are met. When the tolerance parameters are met as shown in step 309, the invention proceeds to calculate the economic effects of the disaster, step 310. In this example, economic effects may influence behavioral factors. Iteration between the economic calculator, as shown in step 310, and the behavioral effects calculator, as shown in step 308, may continue until a user defined tolerance parameter is met, as shown in step 311. In the process of iterating between steps 308 and 310, additional iterations between 305, 306 and 308, according to
After all necessary iterations, steps 307, 309 and 311, between the disaster, infrastructure, behavioral and economic calculators have been performed, the invention proceeds in step 312 to calculate the impact of relevant disaster modified infrastructure, behavioral and economic factors on resource availability. Next, the invention calculates the demand forecast, step 313, which may depend not only on disaster modified infrastructure, behavioral and economic factors, but on resource availability as well. Next, the invention calculates an appropriate allocation of available resources to the forecasted demand, as shown in step 314. This allocation can be performed, for example, by way of a mathematical optimization program, or other algorithm that optimizes one or more objectives of the firm. Finally, the invention computes the impact of the disaster on the business in step 315. Business impact may be measured financially, and/or otherwise (e.g., customer service levels), as may be implied by the ability of the company to complete its value generating operations (e.g., satisfying customer demand for products and services), for example.
To capture the evolution of the business impact time (i.e., over a given planning horizon), there are at least two ways of executing the sequence of steps in
The actual sequencing of calculations may be controlled by a simulation manager, which may also manage the data which is to be shared between calculators. This data may be stored in a database. The purpose of the simulation manager is to facilitate the display of the data on a web-based graphical user interface as well as execution of sub-models with appropriate data access and finally persistence of all the data into the data warehouse for final analysis.
A disease propagation model provides information on how many people in any geographic area are susceptible, exposed or recovered in each time-step (e.g., week or day). Output from the disease propagation model 401 is used as input to an infrastructure model 402, a behavioral model 404, and an economics model 406.
The infrastructure model 402 determines the impact of a pandemic on the business infrastructure, which includes electricity, air transportation, ground transportation, water, and the Internet. The infrastructure model 402 provides a predicted effect that disease propagation may have on the availability of infrastructural elements.
The behavioral model 404 determines the impact of a pandemic and of related mitigation actions on employee absenteeism. The behavioral model 404 thus provides predicted social and psychological effects that disease propagation may have on people.
The economic model 406 determines the impact of a pandemic on the economy, specifically on the gross output for key business sectors. The economic model 406 thus provides a predicted effect that disease propagation may have on key economic indicators.
The final results from the infrastructure, economic and behavioral factors calculators 402, 404, 406 are fed to a demand risk model 454 and a supply risk model 456. The demand risk model 454 determines the impact of a pandemic on the demand for a firm's products and services and may estimate changes in demand by country and by brand. The supply risk model 456 determines the impact of a pandemic on the supply of products and services needed to produce a firm's products and services.
A value chain model 458 estimates the impact of a pandemic on a firm's costs and revenues. The value chain model 458 thus receives input from the demand risk model 454 and the supply risk model 456 and determines an allocation of available resources.
A finance model 459 estimates the realized revenue and cash flow by incorporating the effect of customers delaying or defaulting on payments by different geographies and lines of business. The finance model 459 thus takes the results of the value chain model 458 as input to determine financial indicators over a planning horizon.
A simulation manager 460 provides the pandemic model 400 with data from a data warehouse 470 and controls the sequencing of calculations and the feedback of data. The simulation manager 460 may also receive user input (e.g., mitigation policies, model parameters) and provide user output (e.g., disease maps, revenue, cost, employee availability) via a personal computer 480.
The simulation integrator 510 provides a framework that allows various pandemics sub-models components to plug-in to the solution. This may be accomplished, among other ways, by using the Java programming language to provide interfaces for the sub models to implement and plug-in at the run time. Such interfaces may provide enough information for the sub models to specify input data requirements, execution methodology and output data definition. When initiated, the simulation integrator 510 prepares the data (e.g., loading geography information, disease information, and so forth) and starts executing each sub model. Depending upon the workload, the simulation integrator 510 can spawn new processes to support the execution of additional pandemic models 541 in parallel. The diagram labels the steps that simulation integrator 510 performs in a sequential fashion and list of sub-models that get executed as follows:
The WPS dashboard 530 follows the standard model view controller pattern for accessing, controlling and rendering a view. The dashboard framework communicates with the backend system to access the data using the Java Management Extension (JMX) layer. The simulation manager provides necessary framework to facilitate the exchange of the data between dashboard 530 and simulation integrator 510. The dashboard component, once they receive the data from backend, renders various web pages to display the data or collects the data from a graphical user interface as entered by end users and communicates back to the backend using JMX client APIs. The clear separation between data gathering at the backend and data rendering at the front end facilitates in the development of both such component in parallel there by saving time, cost and promote sharing of such component between dashboard 530 and simulation integrator 510.
The data warehouse 520, which contains both status data and model run time data, is a critical component in the simulation manager architecture. The static data is loaded only once at the start of the corresponding relational data tables as they created in the data warehouse. It consists of data that does not change during a modeling exercise (e.g., geography, airport information, company information, and so forth). The data warehouse 520 contains additional tables to efficiently store the large data generated during each pandemic simulation scenario execution. This data may then be used to perform the post model data analytics to understand the results of each scenario, their impact on end goal and for comparative analysis.
While the invention has been described in terms of its preferred embodiments, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the appended claims.
