Coalition Planning for Operations Other Than War
Dr E. C. T. Walker (Panel Chair)
13 Spring Street
Ipswich, MA 01938
The objective of the Workshop panel on Coalition Planning for Operations Other Than War (OOTW) was to:
The difficulty of obtaining and willingness to share basic data for the planning process from political, military, and civil participants, rather than planning itself, quickly emerged as a differentiating issue for OOTW. Security, policy, doctrinal, and cultural considerations aggravate a basic issue of "organizational will". In addition, the distinct goal of long-term settlement or continuing war avoidance, rather than the goal of finite resolution of a military conflict, distinguishes OOTW. Technical means were discussed for providing an "auction" or "brokering" service and basic forms for presenting and specifying planning information as a means to enable OOTW coalitions to develop trust, share objectives, acknowledge participation, and exchange basic information concerning resource availability and capability during all phases of an OOTW. The panel concluded that a simple brokering service and tools for using it, based on the Internet infrastructure, could be implemented quickly. If successful, this mechanism would aid OOTW planning directly by providing a forum for trust, planning data, and plans to evolve together. Once in place, the service also would facilitate gathering the detailed knowledge necessary for providing feasibility analysis and risk assessment technology specialized for OOTW planning.
The panel initially considered the operational characteristics of a variety of war avoidance operations, including peacekeeping, peace-enforcing, non-combatant evacuation and disaster relief. The working group was fortunate to have several participants with broad experience and expertise in OOTW. The primary source of anecdotes and examples was provided by a detailed, hypothetical "Golden Bowl" scenario made available to the panel by Dr. A. Rathmell of DERA and lessons learned from the IFOR operation in Bosnia, documented by Larry Wentz, the contributing editor of "Lessons from Bosnia: The IFOR Experience."
Among the points discussed were
Some OOTW are directed primarily by the nation affected or a primary participant (e.g., disaster relief), and others by participants experienced at co-operating with each other. Sharing objectives and detailed information about resources is not especially problematic for these relatively homogeneous or practiced coalitions. However, combined operations by partners familiar with one anothers political objectives, resources, and methods of operation contrast strongly with operations conducted by ad-hoc coalitions that involve partners with dissimilar cultures, unfamiliar objectives, doctrine, methods, etc., and undeclared agendas. Such coalitions of heterogeneous organizations were the focus of discussion for the remainder of the Workshop.
Two aspects of such OOTW planning emerged as unique:
While acknowledging the importance of technology for coping with uncertainty, inconsistency, rapid change, security, etc., and providing an information infrastructure for addressing such issues; panel members experienced with OOTW saw developing basic objectives and sharing information about the resources necessary for a planning process to operate as the primary need.
An interesting corollary discussion took place concerning the need for military planners to develop objectives and doctrine for military organizations that are engaged in settlement and war avoidance operations. Such work also is needed to understand how military command and control systems and procedures can be integrated with sophisticated and perhaps preeminent non-military, or even non-government information infrastructures and decision-making procedures. Non-government organizations (NGOs) often are on the ground long before military or government directed coalitions are formed. They will have made intelligence and operational contacts and may have a considerable number of operatives. Some private participants may have unique or uniquely sophisticated capabilities and resources. Generally, todays military organizations and information infrastructures are not set up to interoperate with substantial "uncontrolled" resources or cede authority to other organizations.
The panel concluded that most OOTW are fundamentally collaborative, rather than hierarchical, with much decision-making remaining decentralized, local, and, to some extent, unaccountable. It also emphasized that the real objectives of the participants are likely to remain private (or simply misunderstood) and, thus, block or subvert attempts to obtain specific commitments or assign well-defined roles and responsibilities as prerequisites for planning. In addition, there typically is not much time for preparation, a need for broad participation, and frequent misunderstanding of intent, etc. among participants. Thus, an OOTW coalitions key technical requirement is some means for sharing publicly the data essential for planning to proceed among participants who are reluctant or unable to provide such information.
It is a paradox that the will to plan, the information necessary for planning, and the plan itself initially are unknown and must evolve together, while the objectives, resources, and intent of at least some participants will be concealed. The technical discussion turned to considering technical means to resolve the paradox by enabling coalition participants to convey the objectives of a coalition, elicit reliable input data from members, and construct and execute plans collaboratively.
The basic data required for planning include the
Other Workshop panels discussed at length the difficulties which coalition planners face in developing models for such data and working with the uncertain, conflicting, and changeable data itself. The technical challenges specific to the OOTW come-as-you-are, open operational environment in which information, once made available, is essentially public, and in which participants do not know their partners' private goals and resources include
Corporations also are discovering that a model in which one participant assumes absolute operational authority, directly controls the flow of resources and information, marginalizes uncontrolled participation and otherwise depends on centralized control of participants and activities, is difficult to implement, adapts slowly to change, and is unlikely to be efficient. Even with perfect information, finding a Pareto optimal plan -- in the face of many perspectives for evaluation and many effectors, each of which can affect many other parties and who are not under central control -- is unlikely.
The technical challenge of coalition OOTW is to take such characteristics as given, rather than as barriers that somehow should be circumvented or removed. How can planning technology help parties who don't want to share information discover and negotiate common goals and use resources efficiently, despite cultural impediments and conflicting or vague objectives?
The Internet already is being used as a source and means for exchanging information and as a brokering "marketplace" among parties engaged in a variety of forms of ecommerce without knowing one anothers detailed objectives, plans or resources. In other words, the Internet is providing a means by which the individual pursuit of individual objectives can lead to collective result.
In addition, the Internet is functioning to provide a means for obtaining and validating information through multiple (corroborating or disconfirming) sources and lowered time and access barriers between sources and consumers of information. Models for secure exchange of information to support financial transactions, on-line trading, and consumer e-commerce, as well as technology to support certificate-based control of participation and access are evolving rapidly.
An informed market model of coalition planning, mediated by an Internet style infrastructure, in which potential participants in an OOTW bid for the right to perform specific operational services requested by a "planning coalition," might serve to reduce non-technical barriers to resolving the information sharing paradox referred to earlier. Such auction mechanisms are known to be efficient for allocating resources to meet published needs in the absence of central control. Since examples, models, and mechanisms for implementing such a market exist, an "OOTW Planning Market" might be implemented quickly.
However, one characteristic of an entirely voluntary auction the possibility that there will be no bids for some needs likely would need to be precluded. This vexation could be overcome by bidders of last resort or by an ultimate authority which makes decisions to bias bidding or impose realignment of resources to cover all needs.
An Internet auction offers some obvious advantages to participants in an OOTW
What are the impediments to implementing OOTW brokerage services, assuming the Internet supported auction mechanism itself is feasible? Most obviously, an auction or market requires a practical contracting language for requesting services and making bids, as well as artifacts or documents for collecting data and conveying the structure and content underlying the requests and responses. These two requirements are related the grammar and terminology of the contracting language underlies the physical forms for expressing requests, bids, and agreements, while the forms with their contents constitute expressions in that language.
Familiar, widely available physical artifacts and tools for creating and editing them already exist in the reports and visualizations that common project management tools and their component applications provide. These constitute use tested "forms" which could be implemented as web plug-ins for collecting data to drive an OOTW auction. Examples include
The underlying structure and detailed models of coalition planning are long-term research challenges. However, panel members pointed out that, preliminary structures could be used to augment templates for collecting data and guide their use for gathering basic data from potential coalition participants.
Panel members cited experiments in which even technically simple augmentation of basic forms has proved very useful to real world users. The mechanism contributes to its own improvement in that the initially simple tools provide a means for gathering additional information needed to improve the tools and their use. Such data also is required for research and development to provide yet more sophisticated tools.
The panel members envisaged an evolutionary spiral in which short-term support eventually would lead to a mature simulation-based environment for planning and for doctrine discovery, tool introduction, feasibility analysis, and risk evaluation. In addition to tools for coalition planners and OOTW experts, this environment would serve as a lessons learned repository, training and exercise environment, and source of detailed templates for OOTW plans. Finally, the environment would serve as an OOTW laboratory for long range technical work to develop representations and mechanisms for OOTW.
The technical discussion concluded by noting that there are some features of coalitions, which simplify the challenge of providing technical support for OOTW. First, dual-use or "bridging" functional types can provide a means for integrating otherwise incompatible capabilities. The example of carabinieri providing an interface between police and military participants in IFOR was cited. Second, the many disparate organizations in a coalition have a variety of means for accomplishing their objectives. Consequently, there may be greater probability that an OOTW coalition will discover a viable plan and greater likelihood that situational intelligence and information concerning risk will be available. (One panel member offered a possible theorem that only k parties with agendas that differ by m are needed to ensure that the sum of the information they are willing to share constitutes an accurate picture of the required planning data.)
The OOTW panels discussion period concluded with a proposal for early experiments with technology to facilitate effective coalition OOTW planning and start an evolutionary research process. Two categories of technical support could be made available for experimental use within six months. If effective, such technology could be used operationally within 1-2 years. The following short/medium-term support for OOTW was identified:
These capabilities would support research to create more sophisticated technology that realistically might require 3-5 years to develop and deploy. The following areas for long-term support were identified:
Leverage for experimentation
The panel members offered several examples in which rapid deployment of a rough and ready technical capability has had profound short-term impact on previously intractable operational bottlenecks. The DARPA/Rome Laboratory Planning Initiative provided a number of cases, beginning with the development and use of the DART system in Desert Shield/ Desert Storm. Recent success in augmenting PowerPoint based mission planning templates with timeline editing and temporal reasoning capabilities, and the development and use of the Pacifica scenario constitute additional examples.
The well-documented IFOR experience and the detailed Golden Bowl scenario made available by a panel provide important and significant data and starting points for defining detailed objectives for the experiment.
The panel identified several programs that are investigating planning requirements and developing technology to support planning that already are underway and could be exploited co-operatively (in OOTW style) to provide significant resources. These include
Control of Agent-Based Systems (CoABS)
Among the candidate users and potential participating organizations in the proposed experiment are
Despite repeated attempts by one panel member to include coalition forming and objectives definition as components of the proposed auction mechanism, the discussion focused almost entirely on resource allocation. Thus, one of the major requirements of coalition OOTW and a successful experiment received only cursory consideration.
An early experiment would have to be conducted in conjunction with on-going activities of participants. To be successful, the scenario for the experiment would need to be carefully correlated with those activities. Its details would need to be refined to address the challenges of the participants. The panel also assumed that participants in the experiment would have expert knowledge of their own OOTW processes; the structure of their plans; and of their information systems and databases.
The idea for an experiment was developed during a short brainstorming Workshop. Although the technical and operational discussants have a high level of confidence that the experiment is likely to yield useful operational outcomes and can be mounted quickly with minimal effort, there was not time for a thorough critical discussion or detailed planning.
Conclusions and next steps
The discussions within the OOTW panel were preliminary, but exceptionally concrete. They resulted in a clear proposal for specific collaborative experimentation that could be conducted soon. The next steps involve confirmation from potential participants of their desire to organize and conduct such an experiment and their willingness to contribute the individual effort necessary to plan the activity and secure support. This is a coalition OOTW in action.