HomeAbout UsBrowse This IssueBack IssuesNews DispatchesSubscribing to Army LogisticianWriting for Army LogisticianContact UsLinks

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
Prediction and Cooperation

Editor’s note: This is the fifth and final article on themes for future logistics innovation identified by the Army Logistics Innovation Agency’s Futures Group. The fourth article is titled Energy on Demand.

Of the five themes for future logistics innovation identified by the Army Logistics Innovation Agency (LIA), prediction and cooperation (P&C) is perhaps the broadest. While the other themes—quantum computation and communication, telepresence, designer materials, and energy on demand—have potential to improve logistics effectiveness and operational readiness significantly, P&C can be regarded as the most pervasive and promising of the deep future themes identified by LIA. In an era of accelerating change, particularly with respect to scientific and technological advances, the Army must understand and be prepared to take advantage of the opportunities ahead.

Evolving joint warfighting concepts, such as network-centric operations and sense-and-respond logistics, have laid the conceptual foundation for the articulation and development of capabilities that will allow future generations to capitalize on these efforts. Several Army and joint concept documents address the need for advanced decision-support tools that will permit faster and more effective decisionmaking. For example, the Joint Logistics (Distribution) Joint Integrating Concept (dated 7 February 2006) describes the ability to perform predictive analysis of sustainment requirements in order to enable sense-and-respond logistics.

To ensure mission success, then, the Army must improve its capability to anticipate logistics requirements and forecast solutions. Logisticians need the ability to predict future conditions and environments so that they can proactively mitigate or otherwise eliminate future adverse conditions, such as an ambush of a resupply convoy, poor local weather conditions that reduce visibility, or failure of component parts. Hence, the implications of developing P&C capabilities go far beyond simply reformulating logistics planning factors and optimizing the supply chain and distribution networks. Essentially, P&C is a set of envisioned future capabilities that will allow logisticians to predict future outcomes with a high degree of accuracy and that will provide opportunities for proactive intervention to ensure favorable outcomes of proposed courses of action.

The P&C theme entails a future operational environment in which reduced uncertainty about logistics requirements and course-of-action development increases the warfighter’s freedom of action by minimizing unnecessary constraints in planning or pauses in the execution of missions. P&C is predicated on global connectivity in a network-centric operating environment that will enable logisticians to know, integrate, and synchronize all logistics business processes and expedite decisionmaking. Processes will enable real-time synchronization between resource application and the commander’s intent in support of high operating tempo and distributed operations. P&C will allow logisticians to achieve full control of the supply chain and distribution system to attain desired battlespace effects with negligible delays.

P&C capabilities that are integrated across logistics, intelligence, and operational domains will be fundamental to future sustained combat operations. As a result, logisticians in today’s Army need an understanding of emerging prediction capabilities and their supporting technologies in order to advocate and plan for the introduction of these technologies into the Future Force. Familiarity with the implications of advances in P&C will enable logisticians to communicate requirements to the research, development, test, and evaluation community.

In this article, we discuss P&C with an emphasis on emerging and envisioned prediction capabilities. We also discuss a technology-centric knowledge management framework that will help lead to prediction capabilities of the future.

P&C Overview

Prediction is the act of foretelling based on observation, experience, or scientific reason. General prediction theories and models serve as the basis for drawing inferences from available data. Hence, prediction is based on information, or knowledge, that is used to project future courses of action with a certain degree of accuracy. Cooperation entails seamless, automated translation and communication among organizations, platforms, and digitally equipped agents that enable total interoperability and synchronization among different legacy software systems, networks, and devices.

Together, prediction and cooperation require a globally integrated network that transparently tracks and predicts processes. P&C supports cooperative interaction to sense and record the physical environment while comparing current input features with historical data to derive predictable patterns over time; P&C also provides the capability of acting on those patterns. Entities operating in a P&C-enabled environment will sense and understand contextual meaning, communicate their state and mission, and act to influence the environment.

The ability to locate, identify, and convert data of any kind into required information and vice versa fosters total interoperability and synchronization among disparate software, networks, and devices. This and the extrapolation of given data into the future (using advanced decision-support capabilities) are essential to achieving the following P&C capabilities that will sustain combat operations on a global scale—

  • Project, with a high degree of accuracy, the outcomes of proposed courses of action.
  • Enhance generation of hypotheses and analysis of courses of action.
  • Permit identification of anomalies and atypical patterns.
  • Reduce delays in logistics status reporting and increase the quality and speed of decisionmaking.
  • Assess the effectiveness of specific courses of action.
  • Predict and take action to preempt logistics demand across the entire tactical-to-strategic continuum.
  • Enable accurate predictions of reliability for components and systems.
  • Enable humans and human organizations to cooperate more effectively.
  • Provide seamless, automated translation and communication among organizations, platforms, smart objects, and digitally equipped agents, regardless of data types and contexts.
  • Provide total interoperability and synchronization among different legacy software systems, networks, and devices.

Logistics Implications of P&C

The logistics implications of the envisioned P&C capabilities are profound. P&C will provide logisticians with the ability to know what is happening within all logistics business processes and to synthesize and act on that information. P&C will provide capabilities that permit future commanders to dominate in complex, chaotic, and time-constrained environments. The logistics concept of support will remain fully synchronized with the commander’s intent in support of high operating tempo and distributed operations.

Accurate, timely prediction of unit needs will allow logisticians to properly mobilize and use resources within a short time period to achieve desired battlefield effects with minimum supply system delays. Capabilities that allow prediction and preemption of logistics demand across the entire tactical-to-strategic continuum will change the way the Army, as part of the Joint Force, sustains the fight against future adversaries. Predicting logistics requirements in real-time shortens the decision and planning cycle and, ideally, preempts logistics failures, thereby contributing to a commander’s freedom of action. Areas such as visibility and control of the supply chain, advanced prognostics, enterprise-wide forecasting or forewarning, and advanced planning are the foundations of a P&C capability. A universal P&C capability will help decrease decision cycle time from weeks and days to hours and minutes.

Technology Behind P&C

P&C cuts across multiple scientific and technical areas and requires a multidisciplinary approach. From a sensors and information-fusion perspective, the envisioned P&C capabilities involve the development of advanced integrated sensor systems and knowledge management architectures applied across the entire logistics enterprise. This involves the collection, transmission, fusion and analysis, and exploitation and assessment of real-time logistics data or information throughout the battlespace and global logistics “pipeline.”

From an information technology/command, control, communications, computers, intelligence, surveillance, and reconnaissance (IT/C4ISR) perspective, the focus is on enhanced logistics command and control made possible by the development of advanced decision-support tools.

Predictive methodologies will be applied within the greater context of battlespace awareness, yielding a fully integrated operations, intelligence, and logistics picture. Logistics networks with intelligent agent systems and sensors will support the decisionmaking capabilities commanders need to deal with the risks of increasingly complex variables. (Intelligent agents are software agents that have the ability to adapt and learn.) These cross-domain scientific and technical areas will support the Future Force fighting in a network-centric operating environment involving complex, heterogeneous, and interactive logistics phenomena. Better sensing and interpretation of these logistics data will reduce risk and uncertainty for the commander.

Technology-Centric Knowledge Management

One key to achieving P&C is a knowledge management framework centered on a multifaceted technology infrastructure. Knowledge management refers to the set of processes developed to create, gather, store, transfer, interpret, and apply knowledge. A technology-centric approach to knowledge management focuses on technologies that will enhance knowledge sharing and growth throughout the logistics enterprise. The table at left lists current-day technologies and techniques that support the collection, transmission, fusion and analysis, and exploitation and assessment of this knowledge management framework.

In addition to the existing technologies listed in the table, P&C will require advances in the cognitive sciences; data storage and retrieval; real-time, large-scale, multimodal sensing; the storing and accessing of historical and simulation patterns; on-board prognostics capabilities; physics of failure analyses; and predictive modeling. P&C also will require advances in the use of intelligent agents, integrated sensing and effecting, opportunistic optimization (dynamic reallocation of tasks within subgroups to optimize processes in a more advantageous manner), and automated ontology extraction (automated representation of objects or entities and the relationships among them).

Currently, Government laboratories, academia, and commercial firms are engaged in research efforts that support varying aspects of P&C. A great deal of research is being conducted in the areas of sensors and computational science, using mathematical models on high-performance computers. In addition, cognitive scientific work is progressing toward an understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines. This effort is rooted in several fields, including artificial intelligence, psychology, linguistics, philosophy, robotics, human-machine systems engineering, anthropology, sociology, and neuroscience.

Another area with significant commercial, academic, and Government interest is intelligent systems or agents. Emphasis is on developing a “stimulus response” system that learns during its existence by sensing situations within its environment and learning which appropriate action permits the system to reach its objectives. Efforts are well underway to develop component hardware or software-based systems to assist users by collaborating and interacting on their behalf. Intelligent systems have some commercial viability in manufacturing, supply, and distribution chain operations as well as communications.

As mentioned in the first article of this series, “Quantum Computation and Communication,” (Army Logistician, September–October 2006), quantum computers may enable new classes of models and simulations with processing power and speed that allow for exponentially more accurate predictions of logistics requirements and outcomes. So, just as several scientific and engineering disciplines are experiencing a degree of convergence and overlap, the future logistics themes also demonstrate a certain synergy.

Future Technologies

Advanced sensor nodes will be the starting point for gathering real-time, continuously updated platform logistics data or information. In an operational context, communication among multiple sensor nodes is best achieved using ad hoc wireless networks, which do not require a base infrastructure to support communication among nodes. In the commercial sector, ad hoc wireless networks are not necessarily advantageous since the required necessary infrastructure, such as base stations and routers, are in place. However, for dynamic, evolving situations in which data are obtained from on-the-move sensor networks, ad hoc wireless networks are essential.

In an ad hoc mode, wireless on-board diagnostic sensors (each one representing a “node”) will directly communicate with each other in a peer-to-peer fashion and will not require central access points or routers. Each sensor of an ad hoc network will have wireless communications capability and some level of signal processing and networking of data built in. Common to all ad hoc networks are the capabilities for sensor nodes to collect data, detect the occurrence of events, estimate parameters of the detected event, classify detected objects, and track objects. In addition, the sensor network will disseminate data throughout the nodes of the network as well as to end-users. Individual ad hoc sensor networks of 10,000 to 100,000 nodes, linked by a common communications protocol, will form the components of larger wireless networks.

Low energy, self-organized networks will be required for remote or hostile situations. Self-organized networks will include sensor nodes that can spontaneously create impromptu networks that dynamically adapt to device failures and degradation, manage movement of sensor nodes, and react to changes in tasks and network requirements. Self-organized networks will enable self-aware, self-reconfigurable, and autonomous sensor devices. In an operational context, sensor networks also will consist of sensors of varying types that can be organized into clusters. (See chart below.) The nodes of a cluster will detect locally occurring events, and the cluster will have sufficient processing power to make a decision that can be broadcast to other clusters or a master cluster.

The broadcasting of logistics data or information will require the development of efficient ways to allow the multitude of wireless devices to communicate within the available radio spectrum. Cognitive devices that can figure out which frequencies of the spectrum are quiet and that can negotiate with other devices in the vicinity (a capability called spectrum sharing) will form the larger part of the cluster of sensor nodes. Spectrum sharing is a difficult problem to solve and requires very detailed mathematical models that present the cognitive devices with certain rules. Nevertheless, we expect that cognitive devices will appear within the next 10 to 15 years to alleviate crowded airways.

At the upper echelon, once data are collected, transmitted, and fused, they will be analyzed and used to prepare alternative courses of action. It is important not to lose sight of the fact that the P&C capabilities we discuss in this article will require vast amounts of information across a broad spectrum of domains from physical to cognitive. Bridging these domains requires not only advances in technology but also advances in decision-support tools that synthesize this vast quantity of data and information. Logisticians then will be able to anticipate logistics requirements, enable planning, and allow forecasting of operational solutions that will result in shortened decision cycles, allow for preemptive intervention, and influence mission success. These decision-support tools must incorporate the commander’s intent.

Some advanced decision-support tools currently under investigation (in anticipation of the network- centric Future Force) include Multi-Resource Polymorphic Collaboration, Distributed Smart Enterprise Object Modeling, Joint Battlespace Infosphere, and an intelligent agent-based infrastructure for decision-support systems. All decision-support tools have the following common characteristics: inputs (data); knowledge and expertise to determine what data need to be analyzed; outputs (courses of action); and decisions that ultimately are made by the user.

Key to the overall decision-support process is predictive analytics. Predictive analytics involves the development and use of advanced statistical methods to process data, both current (real time) and historical, in order to make predictions about future events. Essentially, predictive analytics arises from basic prognostic techniques. Today, baseline prognostics techniques are available that allow for data mining and modeling of electronic systems using statistical methods. P&C, however, requires advanced prognostic capabilities not only to detect the early onset of faults (using nonlinear methods) but also to allow the prediction of device failures using the advanced physics of failure models.

Predictive analytics based on future prognostic capabilities will allow for advanced predictive, descriptive, and decision models. Predictive models will analyze data and data patterns to guide decisions; descriptive models will analyze relationships among the many different elements of decisions; and decision models will forecast the results of courses of action. Predictive analytics aids decision logic in order to maximize the desired outcomes of certain courses of action while minimizing other, undesirable outcomes.

Portions of P&C are achievable before 2030 and, with an architectural roadmap, could evolve over time to improve logistics capabilities as new technologies emerge. This would support acquisition of technologies through a spiral development process, giving users exposure to much-needed decision-support tools while providing feedback to developers on desired capabilities and effectiveness measures.

As P&C becomes a reality, its benefits to the warfighter will be significant. It will provide crucial decisionmaking information for operational commanders and enable logisticians to integrate intelligent agents into decision-support tools to control the logistics enterprise. As the Army confronts new enemies and acknowledges an uncertain future, it needs to proactively seek to exploit accelerating scientific and technological change in order to retain battlefield supremacy over our enemies. The Army logistics community must be prepared to articulate clearly the necessary investments in the development of future logistics capabilities in order to support combat systems and forces operating in an ever more complex strategic environment.

Soldiers’ lives and our Nation’s defense depend on the continuous development of predictive tools in order to maintain an exceptionally high state of logistics readiness that does not lag behind other factors of operational readiness. Indeed, because of the nature and complexity of the missions facing operational and tactical forces and the criticality of real-time synchronous integration of logistics with operations, logisticians may need to identify and exploit a wider array of science and technology than do their operational and tactical counterparts.

The intent of this series of articles was to provide members of the logistics community with a preview of future possibilities for Army logistics by describing interrelated areas of basic scientific research as they apply to various logistics functions. Going forward, the Army should realize that its challenges are as much cultural and organizational as they are scientific and technological. Solutions will require collaboration with the research, development, test, and evaluation community as well as with the larger Army and joint concept development and experimentation communities. The Army must ensure that innovative ideas are fully explained and incorporated into the fabric of the Army as it operates as part of the Joint Force. As it adapts to change, the Army can effect far-reaching improvements to logistics processes and readiness.
ALOG

Dr. Keith Aliberti is a research physicist in the Sensors and Electron Devices Directorate at the Army Research Laboratory at Adelphi, Maryland. He currently serves as the laboratory’s liaison officer to the Army Logistics Innovation Agency at Fort Belvoir, Virginia. He has a B.S. degree in physics from Rensselaer Polytechnic Institute and M.S. and Ph.D. degrees from the State University of New York at Albany.

Thomas L. Bruen is a logistics management specialist at the Army Logistics Innovation Agency at Fort Belvoir, Virginia. He has a bachelor’s degree in engineering from the U.S. Military Academy and is a graduate of the Army Management Staff College’s Sustaining Base Leadership Management Program.