Optimizing programs with uncertain funding
In today's highly competitive and uncertain defense appropriations environment, it is critical that programs be prepared for the need to adjust their execution plans in order to accommodate changes in funding from year to year. Defense programs that survive are typically those that are able to maintain momentum even when faced with a declining budget and those that are able to leverage funding increases by turning them into schedule accelerations that ultimately lower total program costs. The need to identify optimal program plans in light of funding considerations is well known; the problem lies in doing so rapidly while securing the concurrence of the program team and customer representatives. This paper sets forth an approach to program replanning in light of funding changes and expenditure variances in defense and government programs. The approach described enables rapid analysis of planning options and the identification of optimal program planning solutions.
When a program experiences a change in funding from that which its initial planning has been based on, it must evaluate how the new level of available resources will best be allocated. The problems with this are twofold:
• First, depending on the level to which expenditures are tracked on the program, there can literally be millions of possible funding scenarios that will achieve the overall program funding allotment. Since a task's schedule and cost are highly dependent on one another, this means that there are literally millions of program plans that could be adopted. Finding those that are optimal is like trying to find a “needle in a haystack.”
• Second, program teams tend to become highly territorial when it comes to the prospect of losing previously planned upon resources or potentially gaining additional resources. Each member of the program team is convinced that they are on the program's critical path and should receive each available program dollar that is available. Securing buy-in by the program's key stakeholders is difficult to achieve.
In light of these issues, a process has been developed that enables programs to analytically identify “optimal” program planning solutions that are compliant with program funding constraints. Programs can optimize on any number of program variables—schedule, cost, risk, etc. Programs might seek to find the resource allocation that minimizes total schedule slip to the program, or that minimize total cost, or that results in the most significant degree of risk reduction. The analytical features of this approach provide a framework that enables the rapid evaluation of the literally millions of potential program plans, and assists in stripping away any perceptions of politically motivated bias among program team members in the allocation process.
The process is based upon the following foundational principles.
1. A task's schedule and cost are highly dependent on the amount of available funding allocated to it.
We know from both experience and intuition that for getting things accomplished on time requires the full portion of resources needed to accomplish the task. When resources are limited programs are less able to accomplish tasks at a desired rate resulting in schedule growth.
2. When a task's funding is cut its schedule slips and the total cost to accomplish the task increases.
Funding cuts almost always result in some sort of schedule slip assuming that the scope of work for a given task is constant. If we've estimated that it will take nine months and $4 million dollars to achieve Preliminary Design of an aircraft structural element, and are subsequently only allocated $2 million dollars, then we will not be finished in nine months. Additionally, it is likely that achieving the Preliminary Design will actually cost slightly more than it had originally been estimated. This is due to the inefficiencies associated with downsizing the team and then attempting to ramp up when the funding constraint is relieved. The cost grows due to the fact that there is a fixed portion of labor that is in place for a longer period of time.
3. When a task's funding is increased the possibility of schedule accelerations and cost savings needs to be explored.
Funding increases may or may not improve a program's cost and schedule picture. Most programs develop execution plans that are as aggressive as possible—since that is what they are rewarded for during the proposal phase of the program. In light of that, many times there are few opportunities to improve on the program's cost and schedule. In such situations funding increases do relatively little to accelerate the program and reduce its cost. Where funding increases do help are in situations where programs have experienced unforeseen difficulties during execution. These increases can help to minimize already realized schedule slips and cost increases. Regardless, the principles are the same as those mentioned previously. Additional resources can enable schedule contractions simply by enabling more rapid progress toward program milestones.
Exhibit 1. Developing Funding/Schedule Relationships
4. These cost/schedule/funding relationships can be mathematically characterized.
Key to employing analytical techniques is the ability to analytically (mathematically) evaluate the funding and schedule relationships associated with a given program task. Mathematical relationships that characterize a task's various milestone accomplishment dates as functions of periodic funding can be developed using regression techniques.
5. Program plans that optimize schedule, cost, or other considerations can be found rapidly using optimization software tools.
Program models can be built that replicate the sequencing of events and logical links present in the program plan using the mathematical relationships for each program task. These models leverage commercially available optimization software tools to explore possible program planning solutions. The model allows for rapid trials of various resource allocation scenarios and monitors the overall program schedule in search of program planning solutions that minimize program schedule slip.
A Four-Step Process
This analytical process to program replanning involves characterizing cost and schedule relationships for each of the program tasks as a mathematical relationships, and then utilizing “off-the-shelf” optimization software tools to evaluate literally millions of possible program plans in search of the optimal solution in light of funding constraints.
Linking Cost and Schedule Information
A program's cost and schedule baseline form the foundation for the analysis. For each program task expenditure profiles are constructed using the baseline cost estimate and plan. From these profiles the expenditures required to achieve various task milestones are identified.
Developing Cost and Schedule Relationships
By linking task accomplishments to funding, a number of schedule results can be evaluated for varying degrees of funding. The relationship between a task's funding and schedule are explored using a software tool that evaluates how milestones are impacted by funding deviations. This evaluation is based on the magnitude of spending necessary to accomplish a milestone, the initial planned time to accomplish the milestone, the achievable resource spend rate, and the degree of “fixed” costs associated with achieving a milestone. Exhibit 1 provides an overview of this process.
Using this process, a series of funding and cost/schedule results is obtained. From this set of results, regression equations can be developed that characterize task accomplishment dates as a function of periodic funding allocation in the following form:
Task Accomplishment Date = A x Period 1 Funding + B x Period 2 Funding +…+ Z x Period n Funding + Constant
(Where A, B,…Z are regression coefficients)
Similar relationships are constructed for each task in the program plan. These relationships enable the rapid evaluation of the impact of funding variations on a given task. A given level of funding on the task can be inputted to the equation and the resulting milestone accomplishment date is automatically determined.
Exhibit 2. Program Optimization Models are constructed in MS Excel
Exhibit 3. Optimization Software Enables Rapid Evaluation Planning Scenarios
Constructing a Program Optimization Model
Optimization software tools are then introduced which allow for the program plan to be modeled using these task relationships. Program optimization models are constructed in Excel (Exhibit 2) enabling the use of off-the-shelf optimization software (Crystal Ball, TK Solver, OptQuest). Logic links and task sequencing rules are modeled to ensure that critical program relationships are maintained between tasks. The model provides insight into the periodic funding allocated to each program task, the corresponding milestone dates for each task, and the residual funding necessary to complete the task. The model is effectively a replica of the program baseline (cost, schedule, and planning logic) built using each task's unique funding/schedule relationship constructed in the previous step. The model has the added capability to establish funding and/or schedule targets that the optimization software will use to solve for as analyses are conducted.
Identifying Optimal Program Solutions
The program optimization model is then used to evaluate program-planning solutions that achieve a given set of program constraints. For example, if the program has a periodic funding cap imposed on it the model can replan the program to ensure that the funding target is achieved in such a way that schedule slip is minimized. The user simply enters the desired program constraint (periodic funding), identifies the program variable that is to be optimized (program schedule), and initiates the analysis. The optimization software automatically explores resource allocation scenarios across all program tasks and seeks to find scenarios that are optimal—all the while working with the constraints imposed by the user and staying true to the planning logic present in the program's execution plan (Exhibit 3). The model runs rapidly, and is able to explore literally thousands of possible program planning solutions in minutes. The model continually identifies solutions and narrows in on optimal areas until an optimal solution is ultimately arrived at. The solution contains resource allocations for each program task, and using the funding/schedule relationships for each task, also contains updated milestone accomplishment dates for each task. This data is used to construct an updated program plan that is compliant with program constraints. The data can easily be exported to a planning tool such as Microsoft Project or FastTrack for rapid updating of the program plan.
A number of applications exist for the program optimization modeling process, including the following.
For most programs funding is a year-to-year proposition. The survival of the program depends on its ability to create and maintain advocacy with Congressional appropriations personnel. Each annual budget cycle brings with it a slew of “what-if” scenarios that are intended to evaluate the impacts of various funding allocations. Program optimization models enable rapid, high-confidence responses to such requests, and enable the identification of program planning solutions that can aid in securing the program's full budget request.
Mitigating Cost Overruns
As programs are in the midst of execution, they are engaged in the constant management of cost and schedule variances. Many of these variances are relatively minor and can be managed without wholesale program changes. There are however those that are so significant that they affect the entire program. In such cases, a program optimization model can be used to identify optimal program replans that seek to minimize the cost and schedule impacts to the program. The model can be used to reallocate resources from other program tasks in such a way as to mitigate the impact of the offending task while maintain program continuity and progress in other areas.
Institution of Cost Reduction Initiatives
It's rare to find a program that is not doing all that it can to reduce its costs. Sometimes the most significant cost reductions are found not in changing the content of a program, but in planning the program's events in an optimal manner. Program optimization models enable programs to explore the breadth of possible program planning solutions to identify those that minimize total cost and schedule.
Proposal Development Activities
Program optimization models provide programs with a replanning capability that is rapid, exhaustive, detailed, and that provides high-confidence program planning solutions. In light of today's funding realities, most RFPs have a requirement in them that teams demonstrate the capability to react to funding or expenditure variances rapidly. Program optimization models are a significant discriminator over the competition when it comes to Management Volume preparation.
The Benefits of the Program Optimization Process
This analytical approach to program replanning exercises has a number of benefits over traditional replanning methods. In the absence of analytical methods, program planning and cost analysis personnel must exercise their best judgment when seeking to restructure the program. While this judgment is sound, the limitations associated with the sheer complexity and enormity of the task make finding optimal solutions highly unlikely. Program optimization modeling features the following benefits.
Program optimization models merge the depth of analytical techniques with the speed of software resulting in the ability to rapidly explore program-planning options in minutes rather than days. Literally thousands of possible solutions can be explored in minutes when using a program optimization model.
Breadth of Exploration
Because program optimization models leverage optimization software tools, there is virtually no limit to the number of possible program planning solutions that can be explored. Manual efforts require a “hunt-and-peck” approach with the hope that the analyst is “hunting-and-pecking” in the right spot. Program optimization models are able to explore the gamut of possible solutions rapidly.
Depth of Analysis
When attempting to assess the impact of funding or expenditure variances the level at which the analysis is to be done drives the amount of effort required. Analyzing the impact at WBS level 2 is obviously less intensive that a WBS level 3 analysis would be. The deeper one is into the WBS the greater the number of tasks that have to be balanced and evaluated. Manual efforts become highly limited at WBS level 3 and beyond because there are simply too many variables to work with. Program optimization models have no analytical limit—as many tasks as can be characterized can be analyzed— enabling far more detailed insight into the impacts of funding and cost variances on the program.
Because the program optimization process can perform an exhaustive and detailed analysis of program replanning alternatives, it nets high-confidence solutions. Because the process leverages the data already documented in the program plan, the results obtained from the analyses are far easier to reconcile with the program's task leads resulting in the ability to more readily secure buy-in by the program's key stakeholders.
Program optimization models save a tremendous amount of time and money. Responding to “what-if” scenarios can be done in minutes rather than days or weeks, and can be conducted without the need to engage the program's task leads in a time-consuming evaluation of planning alternatives.
Program Optimization in Action
The Space Based Laser Integrated Flight Experiment (SBL IFX) program spanned nearly 12 years and was dependent on incremental funding provisions. With each new congressional budget cycle, the program conducted numerous “what-if” scenarios that sought to characterize the likely system deployment date associated with various annual funding amounts. Using the process described here, a program planning model was developed that enabled the program planning staff to evaluate the impact of various funding scenarios in a matter of minutes rather than days, while at the same time identifying higher confidence solutions that were the product of an exhaustive analysis of possible program solutions. Additionally, the model was used in the proposal development process to establish program cost and schedule targets for program team members. The customer mandated proposals for each program increment, yet the funding for the increment was typically very fluid. Attempting to construct a proposal that was compliant with the allowable funding proved to be a very difficult task due to this uncertainty. In light of this, the program used a program optimization model to identify optimal resource allocations across all program tasks for any given incremental funding level. These resource allocations (labor, material, subcontract dollars) were then flowed to the various task leads and served as the primary inputs into their individual BOEs. The program optimization model streamlined the proposal development process and significantly relieved the burden typically imposed on task leads during the proposal process.
For more information on the Program Optimization process visit www.programoptimization.com, or contact:
Attn: Tom Woods
Proceedings of the Project Management Institute Annual Seminars & Symposium
October 3–10, 2002 • San Antonio, Texas, USA