Optimal Scheduling of Resources (Automated Prescriptive Analytics)

Transcription

1 Optimal Scheduling of Resources (Automated Prescriptive Analytics) NBS Enterprises Competition Sensitive Contracts Natasha J. Schebella/ CEO & Owner Administration Tiffany Schebella/ Office Administrator Technical Gary S. Schebella/ Chief Scientist Committed to Our Clients, Our Customers, and Our Employees

2 Optimal Scheduling of Resource (Automated Prescriptive Analytics) INTRODUCTION (NOTE: ALTHOUGH THE EMPHASIS OF THIS PAPER IS HEALTH CARE, THE PARADIGM AND EXISTING OPTIMIZATION ALGORITHMS ARE APPLICABLE TO ANY DOMAIN OF RESOURCE MANAGEMENT) The Department of Veterans Affairs (VA) is faced with the escalation of warriors in need of regular and special health care. The Iraq and Afghanistan conflicts have produced several veterans whose records must be processed with a further need to schedule care within the context of limited resources. To minimize the impact of these issues, the VA has initiated the use of mission planning software tools and the establishment of systems analysis The efforts comprise management cost accounting, health information, patient characteristics and resource availability. Regardless, scheduling has become inefficient and many veterans suffer extensive delays when requesting health care. While responding to asynchronous change and evolving requirements, disparate data sources are in a constant state of flux. Because of system dynamics, additional techniques are required to provide management courses of action while scheduling resources and services in a timely manner. In response to scheduling requirements, NBS Enterprises (NBS) has developed a set of algorithms and a prescriptive analytics system (PA) that associates patient requests and available resources and provides recommended scheduling profiles in minutes, as opposed to hours, days or weeks of manual efforts. The NBS PA system assists with the management of responsive actions that guide physical control. Implementation provides analysis and actionable intelligence: a) Automatic mapping of patient requests to the PA (Metadata) b) Consideration of constraints: priorities, locations, resource availabilities c) Association of requests and resources d) Rapid development of a scheduling profile for a specified time interval - 2 -

3 e) Response to rapid requests and asynchronous events f) Assessment of business models, care concepts and technologies g) Evaluations and prioritization of proposed incremental builds for the VA health care enterprise NBS CAPABILITIES NBS Enterprises (NBS) is foremost in the development of management decision aids. We are able to associate dashboard data from disparate sources such as databases for patient records, personnel, and resource specifics and availability. For example, if a new technology or procedure is required as an addition to veteran s care and perhaps no additional funding is available, a manager will be interested in the impact of the addition upon schedule, priorities and risk. Answers to this query, by a staff operating without automated decision support, require several days. Furthermore, scheduling of services cannot be achieved in a static environment. Rather, value analysis of the VA health care enterprise must be correlated with optimal scheduling software so that enhancements to overall performance are continually visible to managers and their planning staff. The NBS PA system associates dashboards and provides planning guidelines, optimal scheduling of services, and analytical forecasting. Because NBS algorithms and software are fully developed, their applications produce great savings in costs and times to produce enterprise enhancements without the need for research and development funding. IMPLEMENTATION PROCESS The implementation of an automated planning tool comprises three phases: a) Phase 1: Assess current scheduling procedures, define transitions necessary for automation, and develop a rapid prototype for demonstration and proof-of concept purposes. b) Phase 2: Incorporate VA desires and enhancements into the prototype and provide a Beta site for the PA. c) Phase 3: Implement and insert the to-be automated PA into all control centers

4 Using the three-phase approach, the VA will be able to produce patient schedules in a well regulated manner. The intent of this proposal is to address Phase 1 only. Phase 2 and Phase 3 are dependent upon the delivery of a useful prototype. Project Tasks for Implementation Phase 1 Define/ Set the context and objectives of patient scheduling - Review enterprise documents, acquire information, review and evaluate current health care processes, interview personnel and contractors to acquire an understanding of the current scheduling environment. Measure/ Baseline performance - Using the information acquired in the previous task, develop analytical models that represent the current scheduling environment. Demonstrate the models to VA personnel. Analyze/ Use data and mathematical methods to understand causes and effects- Populate the models with relevant variables and parameters. Exercise the models to compute overall mission performance, estimate the impact of new concepts and technologies, and provide requirements analysis and analytical forecasting. Improve/ Develop modifications -1)-Using the information acquired during prior tasks, define a to-be PA/ scheduling tool that responds to evolving requirements. 2) Define the data streams to be acquired from databases, provide metadata (data about data) for the transition of data to model inputs, identify databases, select or develop metadata processors, structure a graphical user interface, a query capability, and a report generator. 3) Refine general algorithms for mathematical analysis. 4) Integrate and test all system components. 5) Design and demonstrate a prototype that automates operations and minimizes the times to produce schedules. Control/ Establish plans and procedures-develop prescriptive enhancements from the results of mathematical analysis. Phase 2 Implementation - After extensive prototype testing, implement a Beta site for further testing and evaluation. Phase 3 Operations - Maintain, manage and operate the PA system

5 TECHNICAL DISCUSSIONS Committed to Our Clients, Our Customers, and Our Employees The approach to the development of a planning tool involves the mapping of data to quantitative, analytical models. The models represent functions, assets and components of a system and act as a guide for an eventual system design. Associations In order for a PA system to be implemented, resources are selected in response to patient requests, metrics and measurements. Requests and resource data are operated upon by transforms and associations are generated for all systems objects. The associated data are them sent to a computational model which computes optimal schedules. DRAFT System Associations DRAFT Requests Plan and Prioritize Transforms Resources Associations Constraints Communications Network Exploitation Humans Algorithms System Processing and Data Flow Data streams exist in various forms such as voice, text, video, and digital formats. Each data source is transformed into a common representation scheme. Relevant data are extracted to form a separate, structured database operated on by unique metadata. A set of associations is developed for all disparate data structures. Dependencies with other associations are identified and are inserted into a representation of a computational model. Based upon the mapping of qualitative data to a quantitative model, algorithms are initiated which replicate all of the statistical analysis of a PA process. The NBS software tool suite is labeled Time Performance 4-5 -

6 Cost (TPC). Thereafter, schedules are ranked relative to their potential for effectiveness. The instantiation of the flow requires a description of all operational tasks and resources. TPC: General Purpose Problem Solver The Time Performance Cost (TPC) tool suite includes a general purpose problem solver that employs one network representation to permit performance computations, optimization and value analysis. A Petri net, which is the primary representation of the tool suite, is a rule-based encapsulation of a network or any complex system. Optimization and variable balancing are always accomplished in the context of a systems model. Only one measure of system effectiveness can be optimized while all system variables are balanced to best achieve an objective. The variables represent competing measures of performance such as minimum communication of messages versus load balancing of software to maximize the processing capability of a network. Either load balancing or message routing can be emphasized, or combinations of the two can be obtained

7 The motivation to use Petri nets as a prime representation scheme results from their power to act as a rule-based system and to be transferred easily into an analytical model or a simulation language. An evolution is made from a system architecture to a Petri net and finally to a model that computes performance statistics. Models are tested by developing challenge cases and comparing automated solutions to those generated manually. After thorough testing, the models are used for real-time computations in dynamic environments. The Petri net backbone is then expanded by the super positioning of neural nets and Bayesian nets. One representation network has the potential to step through the triad, producing performance statistics, optimization and value analysis (isolation of components which produce the greatest influence on system performance). The back propagation technique for neural nets is used for pattern recognition and other learning schemes. The incrementally changed weights associated with network links typically represent correlation statistics. For the composite TPC nets, the weights in addition, signify performance statistics generated by a Petri net computation. For example, if an analyst is optimizing a communications flow, the weights indicate the quantities or latency by message type that are transmitted through each link and node. Back propagation is conducted in response to performance calculations until a best solution for message routing is achieved. An objective function might be the total latency of all messages in a network. The TPC approach is to transition Bayesian nets to Petri nets, which are a directed graph comprising conditions, transitions/ events and connecting links. By superimposing Bayesian nets onto Petri nets, the representation becomes time related. The computational results provide a worth computation of an increment of knowledge

8 . Scheduling Algorithms After data are structured and fed to the TPC general purpose problem solver, a set of prescriptive analytics algorithms are implemented. Decision trees are formed which represent the logic of resource allocations. Thereafter, statistical analysis is performed producing a set of schedules which enhance service effectiveness while reducing time and costs. VA analysts are presented with a prioritized set of suggested actions. Decision are made after review by an evaluation team including queries and what if analysis

9 Value Analysis To fully understand the impact of operational functions and the dynamics of an implemented or proposed system, modeling and simulation are necessary to compute performance statistics for all of their artifacts. Further, the impacts of current and forecasted technologies need to be assessed. Numerous tools and simulation languages are available for applications. However, a COTS tool which encapsulates the capabilities to generate both performance calculations and the optimal disposition of system components is not available on the open market. The NBS paradigm and software tool suite not only derives performance statistics and optimizes the disposition of resources, but it also provides analytical forecasting and prescriptive analysis. Applications of the NBS approach produce the answers to system administrator questions, namely, 1) How well does the enterprise, its systems and its components function (performance analysis)? ; 2) How can we improve operations (optimization)? ; 3) What is the value of a new technology ( component performance in the context of the enterprise)?,4) What is the risk of a program change?, 5) what are overall risk factors and how can they be mitigated?, and 6), How do we plan for the future (analytical forecasting and the transition from as-is to to-be)?. In essence, qualitative representations are transferred to quantitative models, expediting analysis and producing a rationale for change. The NBS approach, which has been shown to be - 9 -

10 accurate and effective, enhances a health care enterprise in an innovative manner, not evident in current management systems tools. NBS applications act as a complete guide for system and enterprise changes while displaying guidance for management decisions. DELIVERABLES The following items will be delivered during the course of the effort: Phase 1 Phase 2 Phase 3 Models of current and to-be scheduling environments. Analytical results of simulations, modeling and optimizations. Performance and requirements analysis, as well as technology assessments. Design of an automated prescriptive analytics tool. Prototype of the automated tool. Tests and implementation of an operational system at a specified location. Management and maintenance of a comprehensive prescriptive analytics tool applicable to all location environments. SCHEDULE Phase 1 will require a three-month effort to develop and demonstrate prototype: Initial analysis and prototype design- first month Prototype development-second month. Test and evaluations-third month The schedules for Phase 2 and Phase 3 will be defined after prototype acceptance by the VA. COST The cost for Phase 1 is $. Costs for Phase 2 and Phase 3 will be defined after prototype acceptance by the VA

11 Appendix A: NBS Corporate Background economically disadvantaged, woman-owned small business EC-WOSB SBA Certified 8(m) enterprise resource optimization capabilities overview We bring together the right combination of expertise and analytics to achieve client-unique enterprise optimization. We architect solutions to complex, cost-driving challenges by leveraging our customer s investments with our analytical techniques, lessons learned, and industry best practices. We employ a proven delivery model that provides skilled industry experts working in partnership with our clients in the employment of our proprietary tools, techniques, and algorithms. We enable our clients to capture a sustainable operational advantage from their investments in... People, Processes, Information, and Technology by breaking through organizational barriers. We empower our clients to act with a fresh perspective. Gary Schebella Chief Scientist. NBS Enterprise LLC We Are Committed to: Our Clients, Our Customers, and Our Candidates 12 1/2 South King Street, Leesburg, Virginia Website: NBS Philosophy and Expertise Political and positional success is directly tied to leadership s ability to lead their organizations, to identify and capture 'at hand' efficiencies, make sound/pragmatic resources optimization decisions, and then continually drive tangible 'continuous process improvements'. Both novel and noble objectives, which are absolutely essential in today s economic reality

12 NBS Enterprises has the right Enterprise Resource Optimization (ERO) model and enabling algorithmic toolkit to empower senior officers with the insight necessary to identify and execute the best course of action for their organization. The NBS tool suite is known as Time-Lives-Cost (TLC). We are ready to help senior officers who are addressing a mandate to capture cost savings and efficiencies across their operations. History NBS Enterprises (NBS) is a woman owned small business founded in 2005 by Natasha Schebella. We have capabilities and experience in three major areas: 1) Decision Support, 2) Information Technology (IT) Solutions and Services and 3) Staffing Services. NBS has personnel that routinely deliver in adverse environments both in the US and abroad, all focused on serving the same mission: the advancement and security of the United States. NBS possesses exceptional talent in the management and technical administration of global requirements for the Department of Defense and numerous commercial businesses. The experience that NBS has gained through programs with customers include the Defense Threat Reduction Agency (DTRA), the Office of Naval Research (ONR), the Department of State and the Diplomatic Telecommunications Service Program Office (DTS-PO), and the Washington Metro Transportation Services, which enables us to skillfully and efficiently coordinate and execute complex planning and decision support. Decision Support NBS has developed logistics decision support algorithms for the Marine Corps, Navy and the Coast Guard. We have also developed optimal assignments for Hospice organizations, matching personnel and resources with patient requests. Based upon our understanding of logistics distribution and the routing of transportation aircraft, we have developed a full compendium of algorithms that provide courses of action for management, communications systems, and the design of sensor exploitation systems. We have also conducted studies and analysis of weapons of mass destruction and have devised exploitation systems that assist in the denial of terrorist threats. IT Solutions and Services NBS has been serving the Intelligence, DoD and Civilian Communities for 9 years. We have successfully delivered solutions in the areas of Information Assurance (IA), Enterprise Architecture, IT Transformation and Modernization, ITIL, Risk Management, Data Center transformation modernization, Security Assessments and Design, cloud environments, Human Resources, and Logistics and Mission support. We specialize in delivering exceptional results on projects with our trained, certified and cleared staff in the most widely used technologies

13 Staffing Services NBS provides direct staffing, temporary to permanent staffing, and temporary staffing in the intelligence, DoD and civilian communities, specializing in information technology and information assurance. The personnel include individuals with top-level clearances and experience. NBS also provides skilled individuals for accounting and financial services, as wells as program and administrative support. We have a reputation for responding rapidly to meet our customer needs. Appendix B: NBS Experience Cost Savings and Performance Enhancement NBS Enterprises has applied their expertise in value analysis and predictive analytics to the development of numerous complex systems. Primary emphases have been sensor exploitation systems, software modeling, logistics, defeat of terrorists, information technology systems, and battle management. In many instances great cost savings were realized. Due to NBS analysis, a reduction of over $20 million was obtained. The Reserve Component Automation System (RCAS) was modeled in entirety, causing the program to be delayed until a new design for both hardware and software is now in production. Because of the new start, several millions of dollars in cost savings has resulted. Similarly, NBS analysis showed that the Comanche Helicopter design was faulty. The program was cancelled saving the Federal Government over one-billion dollars. For logistics programs, cost were not reduced, but performance was enhanced. These programs comprise the US Coast Guard Deepwater, Navy/ Marine Corps distribution of supplies to expeditionary forces, and commercial supply chain management. Performance enhancements were also provided for Space Station Software and the NASA Financial Management System. NBS Projects NBS has participated in the assessment and architectural analysis for the following: Network Analysis and the Optimization of an Information Technology System Design/ FBI Comanche Helicopter Software/ US Air Force NASA Financial Management System/ NASA Reserve Component Automation System/ National Guard and Reserve Saudi Air Defense System/ Saudi Arabia Logistics Software/ US Navy and Marine Corps Space Station Software/ NASA Sensor Exploitation System and Software for Terrorist Defeat/ DTRA Algorithms to optimally distribute assets in response to sensor exploitation system detections/ DTRA

14 Mapped representations of the control system and functional components of the human body to a set of graphs which can be processed as a dynamic net/ Basic Research. Algorithms to optimally select and position software and hardware within a distributed information system network/ Comanche Helicopter. Optimization algorithms for the distribution of Navy/ Marine Corps supplies: synthesis of supplies in response to missions and rapid requests, optimal routing, optimal loading, merging of supply orders, multiple stops per delivery platform and retrograde/ ONR and Dahlgren NSWC Analysis of Sensor Exploitation Systems for the Defeat of Nuclear Terrorists: Devised a systems architecture for the detection and interdiction of nuclear terrorists/ Defense Threat Reduction Agency (DTRA). Nuclear Terrorism, An Assessment of Land Deterrence Concepts: Defined and assessed nuclear terrorist tactics and US countermeasures for an encompassing set of land threats/ DTRA. An Analysis of Sensor Applications and Mixed Deployment Strategies for Nuclear Threat Reduction: Simulated, modeled and assessed combinations of decoys and fixed and mobile sensors to best detect and interdict nuclear terrorists/ DTRA. An Analysis of Biological Detection Systems: Evaluated the biological sensor deployment schemes for Camp Lejeune, NC and San Diego, CA. Provided optimal sensor selections and location schemes for both areas of deployment/ DTRA. Assessment of Delivery Vehicles for Logistics Distribution, Dahlgren Naval System Warfare Center (NSWC): Assessed distribution performance of gliders, unattended air vehicles, helicopters, cargo aircraft, and VTOL aircraft Logistics Workshop for ONR, Documented sense and respond tactics Sea Base Concepts of Operation and Logistics Technology Applications: Modeled and assessed the Navy/ Marine Corps supply chain for ONR. Logistics Distribution from the Sea Base: Modeled and assessed logistics distribution from the Sea Base to expeditionary forces/ ONR. Investment Strategies for Logistics Research and Development: Developed an investment strategy for ONR logistics research and development. A Vision of Focused Logistics for Sea Base Applications: Initiated research and prepared a concept of operations for Sea Base logistics/ ONR. Decision Support for Logistics Distribution: Conducted research and developed analytical algorithms for logistics decision support and course of action computations/ ONR. An Analysis of US Coast Guard Logistics/ Deepwater Project: Assessed distribution procedures and command and control of Coast Guard logistics distribution. Concept Analysis of Naval Logistics Systems: Investigated essentially all relevant technologies and operational procedures for logistics distributions to expeditionary forces/ ON Information System Infrastructure for Naval/ Marine Corps Logistics: Documented descriptions of existing equipment and procedures and recommended a spectrum of potential improvements to the Navy/ Marine Corps information system enterprise/ ONR and Dahlgren NSWC. Marine Corps Fuel Distribution: Investigated and assessed existing fuel distribution system and recommended improvements/ ONR and Dahlgren NSWC