Maintenance, Retrograde, and Waste
Maintenance, retrograde, and waste are not major logistics demands from a tonnage perspective. Yet they are necessary parts of the overall logistics picture, and improvements in any of these areas will improve the logistics system overall. Currently, Army ground vehicle fleet maintenance is generally performed on a schedule (aside from emergencies). In the Army aviation fleet, however, maintenance is increasingly being performed based on condition, resulting in overall efficiency improvements in maintenance. Also, additive manufacturing, known colloquially as three-dimensional printing, is being used in limited contexts in the field and is a technology that promises to improve, and perhaps transform, maintenance. Retrograde comprises not only removing materiel from a theater at the conclusion of operations; it also entails the shipping of reparable parts back to depots for repair, in preparation for those parts to be placed back into the supply chain. This is a critical logistics function, and this report has some suggestions for improving its efficiency. Finally, waste is generated in the course of operations. This waste must either be disposed of on-site, typically in open burn pits and incinerators (with considerable potentially adverse environmental and health impacts), or trucked off-site for disposal. Reducing the waste generated in the course of operations, including that which must be trucked off-site, will lessen this demand on the logistics system. This reduction can be accomplished in a variety of ways. One is reducing waste in packaging. Another is converting the waste to usable energy.
One way to improve maintenance is by conducting maintenance based on condition rather than on schedule, called condition-based maintenance (CBM). There are efforts under way to manage maintenance information, known as CBM+. Another way to reduce the maintenance demand could be to improve visibility into the logistics systems that track ordered parts. This is discussed more in Chapter 6. Finally, additive manufacturing (e.g., three-dimensional (3-D) printing, stereolithography, and selective laser melting) is attracting a great deal of interest as a way to meet demand at the point of need. As will be discussed below, additive manufacturing (of which 3D printing is a subset) has some promise, but it also has some drawbacks and may not be the silver bullet some believe it to be. While additive manufacturing might help alleviate one logistics demand, it will create other logistics demands to support it. It requires energy and, thus, fuel, and it must have raw materials to “print” parts.
The maintenance and repair of systems is a multibillion-dollar annual cost to the Department of Defense (DoD). In fiscal year (FY) 2011, maintenance accounted for $79.5 billion, or 12 percent of the total DoD resource allocation of $689 billion (DoD, 2012). As DoD fields new systems that incorporate innovative and increasingly complex weapons system and platform technologies, and as operational imperatives shift, maintenance and repair sustainment challenges can be expected to increase. For instance, as DoD attempts to reduce operational costs by reducing personnel, the maintenance and repair burden on the remaining personnel will increase. In addition, the aging weapons systems in the U.S. inventory will increase the pressure on the supply and maintenance communities to maintain asset readiness. Platforms are being used well beyond their intended design service lives, and this trend is not projected to change in the foreseeable future. Also, U.S. forces operate in extremely stressing
environments that bring on a high need for maintenance. This stress has led to numerous supply and maintenance issues, including challenges with parts obsolescence, supply chain management and technical data management, wear and corrosion control, component reliability, and test and repair infrastructure. With these increased pressures on the maintenance community, there is a need for technology solutions that will enable the operational improvements desired while reducing the maintenance cost burden on the Services. These solutions would involve enhancements to maintenance technologies as well as to DoD logistics systems.
Over the last 20 or so years, DoD has supported modest programs (e.g., the ManTech Program1) that have demonstrated that platform sustainment costs can be dramatically reduced through the implementation of advanced technologies developed to address platform maintenance challenges. ManTech Program organizations working closely with the maintenance community have developed and transitioned repair technologies that reduce both the cost and time of maintenance activities for specific system components as well as increase the mean time between replacements.
A benefit of deploying improved maintenance technologies would be the reduction in logistics requirements for system maintenance. The ManTech Program has addressed needs in both advanced, depot-level process fabrication technologies and more efficient repair and maintenance procedures.
The need to repair, remanufacture, or reconfigure components for weapon systems represents a supply chain challenge for the defense industrial base. Aging systems and platforms and the related challenge of parts obsolescence will also impact the Services’ ability to maintain fielded systems in the future. Repair parts for older systems may no longer be available. Depots and logistics centers cannot stock sufficient spare components for all of these assets for an indefinite lifetime, so that obtaining them may result in long lead times and high costs. Field-level maintenance is thus constrained by parts availability.
Additive manufacturing is a rapidly developing technology that can support a wide range of commercial and military applications and could support some of the Army’s logistics needs. In additive manufacturing, volumes of material are either added to selectively restore the dimensions and features of a part as a repair or are used to directly reproduce a part via a digital representation of it through a computer-assisted design (CAD) file or a point cloud.2 In contrast to conventional formative and subtractive manufacturing processes, all additive manufacturing technologies fabricate features or components in an additive manner through the layer-wise addition of material. The desired component dimensions, or shape, of the part are achieved through the coordinated motion of a heat source and the material feedstock to repeatedly produce layers of fused material. Objects produced by additive manufacturing usually require some finishing before they are ready for use. Accordingly, many additive manufacturing facilities have three- to five-axis mills for finishing the parts. The combination of additive manufacturing and on-site finishing capability also supports the ability to engage in rapid prototyping.
There is a tremendous level of public and private sector investment being made to advance the technology and to commercialize products made by additive manufacturing. Volumes of polymer-based material ranging from a 1 mm3 repair to a 1 m3 component can be produced with this technology. At present, polymer-based additive manufacturing materials and processes have found many application areas in industry, and products from these processes are being used in a variety of applications. Metal-based additive manufacturing is less mature and is still an area of active research for material and process development and for process control and optimization development—for example, under the Defense Advanced Research Projects Agency’s (DARPA’s) Open Manufacturing Program and the additive manufacturing-focused, public-private partnership, America Makes. Fundamental research is ongoing in
________________________
1 Additional information about the DoD ManTech program is available at https://www.dodmantech.com/.
2 A point cloud is a collection of points in a coordinate system that defines an object.
other areas that range from multimaterial integrated structures for features such as embedded sensors and electronics to the biological printing of human organs. There are two Network of National Manufacturing Institutes that are funded by DoD. The one at Youngstown, Ohio, funded by the Air Force, was the first such institute and is addressing additive manufacturing in metals for specific parts. The second, which is located in Chicago, is funded by the Army and managed out of Huntsville. It focuses on the digital supply chain, which can have great impact not only on manufacturing but perhaps allow for the robust development of the digital chain for both direct and retrograde logistics. Additive manufacturing has a wide range of possible applications using a wide range of materials.
Additive manufacturing technology is well suited for the repair of high-value components or the production of small lots of components, and it has the potential to address supply chain concerns associated with surge production and long-lead-time items. Thus, additive manufacturing has great potential for addressing the availability of parts and components for critical DoD assets. The ability to provide a repair or a replacement part on demand can bring increased readiness and affordability and could provide surge capacity for sustainment activities of the defense industrial base. In addition, the ability to quickly produce innovative products specifically needed to address emergent operational needs is a unique benefit of the technology.
There have been several demonstrations of repairs using additive manufacturing in DoD. For example, the Anniston Army Depot demonstrated the use of laser-engineered net shaping to repair gas turbine engine components on the M1A1 Abrams tanks (Zhang, 2010; Optomec, 2006). The Applied Research Laboratory at Pennsylvania State University and its partners have demonstrated additive manufacturing repair techniques for titanium compressor blade tips in the F402 engine, valves and shaft components for submarines, gear components for aviation, and aluminum shells used in undersea systems for the Navy. Researchers at Rolls-Royce developed a laser-engineered net shaping repair for high-performance Ti-6Al-4V blisk aerofoils (Tuppen et al., 2006). In addition to depot-level repairs, the Army has pioneered field-level maintenance applications of additive manufacturing. The U.S. Army Tank Automotive Research, Development and Engineering Center has examined the incorporation of additive manufacturing systems into the mobile parts hospitals deployed in Iraq and Afghanistan.
The Army Rapid Equipping Force funded Exponent, Inc., for an expeditionary laboratory support system, which included polymer and metal-based additive manufacturing capabilities.3 This system made some parts for soldiers in the field much more quickly than waiting for manufacture and shipment from the United States. Under a current Industrial Base Innovation Fund program, the U.S. Army Armament Research, Development and Engineering Center is addressing critical issues so as to accelerate the implementation of additive manufacturing technology for sustainment activities both in the Army and in other Services. This includes identifying families of components that can be repaired and remanufactured by additive manufacturing, the establishment of standards for qualifying parts made by additive manufacturing, development of best practices and procedures for quality control, and education of the defense industrial base. In addition, many aerospace original equipment manufacturers are pursuing additive manufacturing technology for the manufacture of new parts and components.
Despite these often positive (although dispersed) technical successes, there are still engineering challenges to the widespread adoption of additive manufacturing as a tool for improving sustainment in the defense industrial base. For instance, additive manufacturing can be energy-intensive relative to conventional manufacturing, with the ratio having been estimated at approximately 100:1 (Choudhury, 2013). The requirement to melt the materials used in this technique is a fundamental aspect of additive manufacturing, and the energy requirements are not likely to lessen significantly over time, even with research. Energy use would also be a significant issue for the forward deployment of additive manufacturing facilities. In addition, additive manufacturing is not a fast process. Based on run times from the additive manufacturing laboratory at North Carolina State University, making a single 8-inch-high titanium part can take from 40 to 120 hours on an ARCAM machine (a Swedish-produced 3-D
________________________
3 Matthew Cox, “Mobile Labs Build On-the-Spot Combat Solutions,” Military.com News, August 17, 2012, http://www.military.com/daily-news/2012/08/17/mobile-labs-build-on-the-spot-combat-solutions.html.
printer capable of printing titanium, stainless steel, and copper articles), depending on the footprint of the part. These build times will improve over time, but the energy cost will probably increase because faster manufacturing will mean melting more material more quickly.
The long run times make a stable electrical source an absolute necessity for field operations, where a disruption in the power supply might cause a partially printed part to become scrap. This would be especially true for hot processes like the ARCAM machine. Cold processes are more forgiving of power interruptions, but objects manufactured by a cold process require the relief of high internal stresses by a heat treatment that involves more equipment, more facilities, and more energy.
Additive manufacturing technology provides tremendous design flexibility in the production of metallic material for structural components. This flexibility can have positive consequences if well understood or potentially negative consequences if not fully understood or controlled. This is because in additive manufacturing, process parameters will directly affect the microstructure and properties of the components produced and, accordingly, their strength and hardness. This aspect of additive manufacturing is the focus of technology programs currently under way to understand the performance limitations of current materials and processes and to come up with improved materials and processes.
A major challenge standing in the way of DoD’s acceptance of additive manufacturing is the lack of methods and guidance for process qualification and component certification for a wide range of metals.4 The issue is ensuring that a part made by additive manufacturing will meet the requirements and standards of the conventionally produced part it replaces. The only existing specification is the aerospace materials specification Titanium Alloy Direct Deposited Products 6Al - 4V Annealed (SAE, 2011).5 This standard considers the deposition of Ti-6Al-4V in terms of testing requirements, minimum properties, and reporting requirements to achieve certification. The American Society for Testing and Materials F42 Committee on Additive Manufacturing Technologies is in the process of developing a range of standards for additive manufacturing processes. This is clearly an issue for Army aircraft, as parts that fly must be certified. Certification requirements will differ based on platform type—for example, a tank has different requirements from an aircraft.
To obtain qualified and certified components, the aerospace industry has conducted a great deal development work on what are called “design allowables” for Ti-6Al-4V processed by additive manufacturing. Additive manufacturing opens up the design space for components by allowing the design of components that are lighter than but just as strong as components that are limited in design by conventional machining. Design allowables would be approved variations in part construction to account for the capabilities of additive manufacturing. Extensive testing to satisfy conventional design and engineering requirements is often required by organizations wishing to implement any new technology for emerging and existing components and platforms. The development of cost-effective certification protocols will be critical to additive manufacturing technology implementation.
Additive manufacturing is a technology that is moving ahead extremely rapidly in terms of new applications. It may be that the best strategy for the Army is to watch very closely the developments coming from industry and to adapt the new applications to the Army’s uses. This applies particularly to the certification of aircraft parts. That said, additive manufacturing is not yet a replacement for conventional manufacturing, particularly high-volume manufacturing. The cost and energy requirements are just too great. It does, however, hold great promise for the repair or remanufacture of parts and components and for very low volume manufacturing.
________________________
4 Chris English, GE Aviation Senior Engineer, “An Overview of Additive Manufacturing at GE Aviation: The Need for Industry Collaboration in Overcoming Barriers,” presentation at the Additive Manufacturing Consortium Kick-off Meeting, Edison Welding Institute, Columbus, Ohio, February 10, 2010.
5 Craig A. Brice, Materials Engineer, Lockheed Martin, “Direct Manufacturing at Lockheed Martin Aeronautics Co.,” presentation at the Additive Manufacturing Consortium Kick-off Meeting, Edison Welding Institute, Columbus, Ohio, February 10, 2010.
Finding 5-1. Additive manufacturing provides a capability for producing components in support of Army logistics system needs at the point of need. Additive manufacturing efforts are ongoing across the Army and are close to the state of the art. However, further technology development is required to fully realize the benefits of additive manufacturing. Owing to its particular energy and materiel demands, additive manufacturing will happen at the depot level for the time being.
Recommendation 5-1. The Army should leverage the industry investments being made in the field and support technology areas that map to the specific needs and implementation barriers of the Army. The Army should support standards development that would form the basis for qualification of components. The Army should work with the other services to address standards for additive manufacturing and certification of parts for procurement.
Finding 5-2. The Rapid Equipping Force’s Expeditionary Additive Manufacturing Laboratory is a solid foundation on which to introduce additive manufacturing capabilities into the Army’s logistics enterprise, as appropriate.
Recommendation 5-2. The Army should continue to support activities initiated by the Rapid Equipping Force to develop a distributed additive manufacturing network that makes use of both organic and commercial capabilities. This network would be utilized to determine the applicability of additive manufacturing to critical Army components as well as to qualify procedures. It would include depots and both academic and industrial laboratories. It could also be a test-bed for integrating field-based maintenance requirements into a distributed design and manufacturing network.
CBM is a capability that evolved from work performed in reliability-centered maintenance (RCM) over the last few decades (SAE, 2002). This contrasts with traditional practices of unscheduled replacement upon failure and flying-hour-based replacement during scheduled, phase maintenance for aviation platforms. These traditional maintenance models tend to presume that component condition is exclusively a function of age and to ignore other important exogenous factors such as environmental conditions, manufacturing variances, preventive maintenance history, and, especially, the different types of missions.
RCM embodies the engineering analysis of the probability and consequences of failure for operating equipment. Such analysis informs system design and streamlines maintenance operations. It leads to a preventative maintenance program designed to minimize the impact and cost of component and system failures. CBM refers to an activity within RCM that senses impending failure and enables corrective action prior to catastrophic failure events. CBM uses sensors, either on the platform or brought to the platform, to assess the health and status of important system components. The seminal work on RCM was based on work performed in the commercial aviation community in the 1960s and 1970s (Nowlan and Hemp, 1978). Since that time, RCM and CBM processes have been developed and implemented by a wide variety of commercial organizations. DoD has been developing and utilizing RCM and CBM tools and methods for many years as a means to improve maintenance operations.
More recently, DoD has become engaged in the development of condition-based maintenance–plus (CBM+), which builds on RCM and CBM by adding the enterprise-level infrastructure required to manage maintenance information. DoD is focused on the development and implementation of CBM+ to improve mission performance and enhance asset visibility for mission planning, effectiveness, and
combat power. CBM+ will provide higher readiness, lower maintenance costs, and improved system safety for DoD assets. Because of the broad impact of CBM+, DoD has established a Joint Service CBM+ Action Group under the Maintenance Executive Steering Committee. Within the Army, the G-4 has the responsibility for CBM+ policy, staff oversight, and monitoring of CBM+ implementation working through the Army Business Process Council. It is well recognized that many stakeholders must be engaged in order for CBM+ to be fully implemented within the Army, and well-structured coordination groups have been working for a number of years on development, demonstration, and implementation of the core components of CBM+ capabilities. Roadmaps and implementation guides that reflect the collective plans and actions of the Army community have been published over the last decade (DA, 2007; DoD, 2008). These documents describe policies, roles and responsibilities, and strategies for CBM+ implementation within the Army.
Coordinated activities are developing the infrastructure needed to implement CBM+ across Army systems. Responsibility for implementation of CBM+ capabilities for specific systems resides with platform program managers. Cost-benefit analyses developed by Army program managers and Program Executive Officers for application of CBM and related topics such as Vehicle System Health Management and CBM+ for specific systems are an important element of DoD’s implementation policy. Cost-benefit analyses conducted by Army commands on implementation of these advanced logistics capabilities have consistently indicated substantial benefits.6
The current status of CBM+ development and implementation in the Army varies by system type (DA, 2007). Army aviation is furthest along in CBM+ implementation. There is strong evidence for the value of CBM+ from the experiences in the Army aviation community. Ground system programs are actively pursuing initial implementation while developing the core infrastructure required for ground systems. The Army missile systems community has a number of CBM+ programs under way. The C4ISR community has programs under way as well; however, systems health management for electronic systems is not yet mature.7
Finding 5-3. Condition-based maintenance–plus (CBM+) supports the goals envisioned in force-multiplying technologies for logistics by enabling the reduction of process costs in the logistics enterprise.
Finding 5-4. CBM+ has the potential to significantly reduce the Army’s logistics expenditures.
Recommendation 5-3. The Army should require the implementation of CBM+ on all future Army major system acquisitions without the possibility of waivers.
Connecting CBM to the Supply Chain
Internal platform-focused prognostic capability, which is the current focus for CBM, needs to be complemented with an understanding of the historical consumption patterns and usage trends associated with operational and environmental demand factors external to the platform. These differential effects of operational mission types (e.g., training, combat, stability operations, and humanitarian support) and environmental conditions (e.g., altitude, temperature, humidity and salinity, sand and dust) can be measured by statistically evaluating the empirical consumption patterns associated with recent deployments. This is the essence of a mission-based forecasting (MBF) initiative, further described in Chapter 6. Research has shown that “demand lead times behave in a fashion that is exactly the opposite of supply lead times. An increase in demand lead time improves system performance exactly like a reduction
________________________
6 Ken Beam, U.S. Army Logistics Innovation Agency, “CBM+ IT Bridging Infrastructure,” presentation to the CBM+ Advisory Group, Washington, D.C., September 25, 2012.
7 C4ISR stands for command, control, communications, computers, intelligence, surveillance, and reconnaissance.
FIGURE 5-1 The prognostic replacement alert signal and order ship time (OST). When OST is less than the alert replacement time, forward (tactical) stocks can be reduced. NOTE: VoS, velocity of supply; TTF, time to failure; RvS, reliability state vector; Tf, time of failure.
SOURCE: Parlier (2013).
in supply lead time” (Gavirneni and Sridhar, 1999, p. 444). This relationship, together with the feed-forward control concept in adaptive control theory, seems to suggest that it is possible to positively affect system output a priori.
Adopting Bayesian updating by creatively combining these two new capabilities, CBM and MBF, holds great promise for improved demand forecasting. The basic logic underlying Bayesian methods is the notion of conditional probability and the systematic incorporation of prior knowledge and expectations about probability distributions into statistical analysis. This method consists of a coherent set of axioms that converts prior information (derived empirically from appropriate historical data using MBF) to posterior evidence (i.e., a revised estimate including new information) by conditioning on observed data (current CBM status).8 Hence, the logic develops an updated forecast in a dynamic environment. Figure 5-1 shows this in graphic form.
For example, MBF can be used to determine well in advance what will likely be required based on a clear understanding of typical consumption patterns for a platform type (e.g., AH-64) performing specific missions (major combat operations, stability operations, noncombat evacuation operations, and so on) under environmental conditions associated with, or similar to, those that prevail at the geographic location where the deployment will occur. Much of the uncertainty associated with external factors in operational demand forecasting will be significantly reduced by using MBF in this way. Then when the
________________________
8 For a theoretical perspective on Bayesian statistics, see McGee (1971). For applications using revised estimates, see Section 15.3, “Revisions of Estimates,” in de Neufville (1990, pp. 279-285), and for practical applications of Bayesian statistics to spare parts forecasting, see Sherbrooke (1992, pp. 71-94).
FIGURE 5-2 Test results for the AH-64 nose gear box. As prognostics improve, significant reductions in forward stocks can be achieved without degrading readiness. NOTE: RMSE, root mean square error.
SOURCE: Parlier (2013).
operation is actually conducted with a designated unit and its particular complement of AH-64s, each with its own internal set of depot-level-reparable life-cycle reliability profiles (for remaining useful life), CBM can be used to revise the original (albeit much improved) a priori estimates provided by MBF.
As CBM matures and both data collection and analytical methods improve, this Bayesian approach is likely to dramatically improve demand forecast accuracy for spares and repair parts (which are Class IX supplies). These empirically derived conceptual advances promise to serve as a similar foundational and methodological guide to improve demand forecast accuracy for other planning domains and classes of supply as well.
The test case for connecting CBM to the supply chain is for an AH-64 nose gearbox (a depot-level reparable, or DLR) and compares the current baseline (using empirical data for this specific DLR) with simulated results obtained by anticipatory ordering of the DLR a number of days prior to needing to replace the part based on the prognostic sensor prediction.
The results in Figure 5-2 shows how readiness (actual cell values expressed as fleetwide nonoperating days, with red, amber, green representing lower, the same, or better readiness) and inventory (expressed as the number of DLRs, the percentage of aggregate stock reduction in the supply system across all echelons, and the associated dollar value reduction) are impacted by adopting this anticipatory replacement policy.
The potential, in the case of just this one DLR, is significant. Readiness can be improved, or inventory can be reduced, or (in the real world) both. The matrix quantifies these various trade-offs. Savings can be a few hundred thousand dollars—again for just this one DLR. When this policy is applied for all (or even most) sensor-equipped DLRs (those with imbedded CBM capability) across fleets—and eventually to ground as well as aviation for the Army—the savings in aggregate stock reduction can be many hundreds of millions of dollars, perhaps even billions of dollars. This saving is achieved because it is no longer necessary to buffer uncertainty at the retail—that is, the unit—level by forward stocking these expensive DLRs (Parlier, 2011).
As noted earlier, DoD supply chain management has been on the Government Accountability Office’s list of high-risk activities across the entire federal government since 1990. The Army, for instance, cannot determine how readiness would be increased by investing more in spare parts. It asserts there is no direct correlation between the level of investment in spare parts and the impact of such investments on system readiness, saying this is due to factors such as maintenance capacity and training requirements (Parlier, 2005; GAO, 2003). The Army needs a plan to overcome critical spare parts shortages (GAO, 2003). Furthermore, it has not been able to conduct coordinated systemic improvements across the multiple organizations involved in the supply chain. This inability has been attributed to the numerous complexities associated with separate, diverse, and independently operating organizations, which are compounded by a lack of accountability and authority for making improvements across the enterprise. Finally, previous transformation efforts have not provided a clear vision to guide, gauge, and synchronize future supply chain improvement efforts by specifying the performance goals, programs, milestones, and resources needed to achieve the stated objectives.
The Army’s inability to relate strategic resource investment inputs to fleetwide readiness has serious consequences for the tactical warfighting Army. The Deputy Under Secretary of Defense responsible for logistics policy stated, “Whether push or pull, our current logistics are reactive. [We have] an industrial age vendor struggling to satisfy an information age customer. Reactive logistics—the old logistics—will never be able to keep up with warfare as we know it” (AUSA, 2004, p. 2).
Finding 5-5. Connecting CBM+ demand information directly to the supply chain could enable advanced scheduling of line reparable unit replacement and preclude replacement before needed. This approach could identify the need to replace a part before it fails. Field testing has demonstrated that such a connection to the inventory system can significantly reduce the requirement for forward stocking of repair parts and dramatically reduce customer (i.e., tactical unit) demand uncertainty.
Recommendation 5-4. As prognostic credibility and accuracy for CBM+ advances, the Army should adopt connecting CBM to the supply chain as inventory management policy, as described above.
The material in this section is drawn mainly from Transforming U.S. Army Logistics: A Strategic “Supply Chain” Approach for Inventory Management, by Greg Parlier, a committee member (Parlier, 2005).
The reverse logistics pipeline, known as retrograde, constitutes the U.S. Army’s value recovery process for reparable spares and includes the vast majority of the value requisitioned at the unit stage. Retrograde constitutes the effort to maintain, repair, overhaul, upgrade, and return large subassemblies and replaceable units that are not consumed but are used as capital assets. Although these items constitute less than 25 percent of the number of demand requisitions, they also represent more than 75 percent of total requisition value. Inefficient retrograde will impose unnecessary costs on the logistics system, will cost time, and will impose logistics burdens associated with the need to transport materiel such as fuel.
From the perspective of systems-control theory, it is important to see retrograde as a feedback loop, one with obvious impacts on output, in this case embodies as unit readiness. Both system dynamics
and control theory suggest the responsiveness of retrograde should have considerable impact upon readiness (i.e., operational availability); it is clearly a feedback loop in the supply chain. For reparable items, retrograde forms a closed-loop supply chain to return, rebuild, redistribute, then reuse (repeatedly) capital assets (e.g., DLRs) to continuously enhance readiness. Operating within the larger logistics structure, this closed-loop supply chain creates internal system behaviors that can potentially be changed through feedback to regulate the output of readiness.
From a practical logistics network perspective, the more DLRs that are delayed in retrograde awaiting evacuation for repair, the more inefficient and unresponsive the reverse pipeline becomes, increasing the overall number of those DLRs required. For every DLR in retrograde, another similar serviceable DLR is needed somewhere in the forward supply pipeline, thereby increasing systemwide demand for that DLR. Or, if the DLR is not available, customer wait time for a back order increases, impacting readiness at the unit.9 This dynamic contributes to the tendency to mass large amounts of supplies forward, the familiar “iron mountains,” placing additional demands on the logistics system. Although not yet well recognized, this is characteristic of a very tightly coupled system that can have potentially drastic negative consequences with little or no warning.
Readiness-Responsive Retrograde
A decade ago, for the first time, the U.S. Army Logistics Support Activity mounted an effort to identify, measure, and quantify delay times for DLRs awaiting retrograde in the reverse pipeline. Total retrograde time average values were found to vary considerably across the various overseas commands but are generally measured as several months. Additionally, there was extreme variability in the numbers of these items returned to various sources of repair (e.g., original equipment manufacturers and depots), ranging from a few hundred to several thousand DLRs per month (Pew Project on National Security, Energy and Climate, 2011).
Subsequent U.S. Army Logistics Support Activity efforts focused on ways to capture reverse logistics information, measure total retrograde times for the various theaters, and estimate the value of reparables delayed in retrograde. However, until recently, there has been no focus on defining and quantifying both the potential reduction in aggregate DLR inventory requirements and the effects of reduced customer wait time on readiness that could be achieved by synchronizing retrograde flow and depot operations with the forward supply chain. For example, the U.S. Transportation Command recently announced improved retrograde operations. But the criterion for this was only the reduction in transportation costs by using more less-expensive surface shipping and less more-expensive air transport (G-4 Public Affairs, 2012). Enormous improvements are possible if retrograde is viewed as a dynamic feedback loop with multiple effects—a closed-loop supply chain rather than independent, disconnected operations with linear additive effects. Recent efforts have quantified key relationships between aggregate inventory size (and investment costs), retrograde velocity (the speed of DLR returns from tactical units to depot facilities for repair and overhaul), and associated transportation costs and their impact on tactical readiness for aviation units. Using actual data for specific DLRs, retrograde relationships were established, along with trade-off curves between inventory cost, retrograde speed, transportation cost, and readiness. This could provide a model for a readiness-responsive retrograde system. Such a model could have enormous potential to both improve readiness and reduce total life-cycle costs.
Finding 5-6. The potential for further improvement in retrograde seems considerable. The various depot-level reparable (DLR) network links and flows, including reverse pipeline flow, depot production and scheduling operations, and forward supply chain flow, must be connected and afforded in-transit
________________________
9 Grace M. Bochenek, Director, U.S. Army TARDEC, “Army Power & Energy: Enhancing Mission Effectiveness, While Preserving Future Choices,” presentation at AUSA Annual Meeting and Exposition, October 11, 2011.
visibility. Then the Army’s extensive investment in DLR assets can be reduced and, through better management within a synchronized, closed-loop supply chain, both current readiness and future capability can be improved.
Recommendation 5-5. The Army should recognize the potential for efficient retrograde operations to enhance unit readiness. It should adopt a new paradigm of readiness-responsive retrograde as discussed above for the crucial closed-loop retrograde supply chain within the larger logistics enterprise.
Improving Retrograde Efficiency
Over the past two decades the Army Materiel Command has incorporated Lean and Six Sigma manufacturing concepts into depot management practices with measurable success in reducing process variances and rebuild times. The focus of the Lean concept is on better synchronizing process flow, reducing work in progress (stagnant inventory) and waste, leading to a just-in-time approach to meet demand. Six Sigma’s complementary focus is on reducing defects to improve product quality by reducing product variation, the proximate cause of product defects within the manufacturing process. Significant additional improvements can now be obtained by adopting synchronized manufacturing (also known as optimized production technology from the theory of constraints, or TOC) for depot repair management. In essence a business-process redesign approach, the TOC enables significant gains in effectiveness (in contrast to efficiency gains) that become possible within a truly synchronized closed-loop supply chain for DLRs. This becomes possible because TOC views the process itself as potentially flawed, an approach that generally aims at identifying weak links in the chain. TOC emphasizes the bottlenecks in the supply chain, improving output by focusing exclusively on these bottlenecks and resolving them. Increasingly, companies that combine their Lean and Six Sigma efforts with a TOC process redesign approach (Caterpillar, Merck, Johnson & Johnson, and IBM, among others) are finding “more success redesigning whole processes,” improving weak links, and reducing or eliminating constraints to improve cost-effectiveness and productivity (ARCIC, 2010).10 It needs to be understood that, like Lean Six Sigma, TOC requires personnel and leaders to be trained as well as ample time to master the skills.
An example of the dramatic improvement that can be obtained using TOC within the military depot system is found at the Marine Corps maintenance facility in Albany, Georgia. Costs, work in progress, and repair cycle times have been reduced, resulting in increased throughput and improved scheduling. The specific results for MK-48 engines have been especially dramatic: Averages and variances for both repair cycle time and labor-hours per engine have been cut in half, with MK-48 engine output per month more than quadrupling.11
Synchronizing Retrograde with Depot Repair
Depot maintenance activities have historically experienced delays in being provided with consumable parts for repair and overhaul. This is due partly to the 85 percent target used by the wholesale system for supply availability. but it is also increasingly due to dwindling numbers of manufacturers of materiel and to issues of obsolescence, especially affecting wiring, avionics, corrosion, and dynamic components. degradation caused by aging aircraft systems and subassemblies. As retrograde efficiency and responsiveness improve, however, the combination of in-transit visibility and emerging health
________________________
10 Vic Ramdass, Operational Energy Office, Headquarters, Department of the Army, “Army Operational Energy Overview, Increasing Mission Effectiveness while Preserving Future Choices,” available at http://netzero.asu.edu/files/vic_ramdass.pdf, accessed October 23, 2014.
11 Grace M. Bochenek, Director, U.S. Army TARDEC, “Army Power & Energy: Enhancing Mission Effectiveness, While Preserving Future Choices,” presentation at AUSA Annual Meeting & Exposition, October 11, 2011.
monitoring systems for weapons systems, including condition-based maintenance, can provide anticipatory information for depot repair before the component actually arrives through the reverse pipeline. Hence, particular or unusual parts can be ordered before rather than after the components and end items arrive for repair, further reducing depot repair turnaround time.
The intelligent collaborative aging aircraft spare parts support (ICAAPS) project, a Logistics Management Institute initiative for the Navy, explored the value of such an anticipatory ability, illuminating the potential for reducing these forecasting lag effects, especially on consumable parts needed in maintenance and repair. Because current projections for depot-level consumable part requirements are based on historical depot repair maintenance data, there is considerable delay between when the parts are actually needed and when theyare incorporated into future requirement projections. The ICAAPS project successfully estimated correlations and relationships between both depot-level consumable parts usage and the operating environments and field maintenance activities that could be expected to affect future depot-level maintenance requirements. This power to more accurately predict parts requirements and ensure they they are available for use in repair when an aircraft arrives rather than at some later time, cut the growing gaps between predicted usage and actual usage by over 50 percent for many consumable parts. Consequently, by expanding the maintenance planning horizon to include all relevant information gathered during the entire operating cycle before an aircraft arrived at the depot for repair, ICAAPS was able to significantly reduce forecasting lag, improve part requirement forecasting accuracy, and reduce depot repair cycle time.12
One of the great challenges to better synchronizing depot repair operations is overcoming the inability to see all of the potentially useful information that could contribute to better forecasting. No integrated knowledge base currently exists to combine aircraft onboard data with potentially relevant unit-level operational information and program depot maintenance data, because each information source is typically maintained by different organizations in multiple, geographically dispersed locations. Recognition of this limitation recently led to a joint initiative between the Air Force’s Oklahoma City Air Logistics Center, which overhauls KC-135 tankers, and the Department of Energy’s Pacific Northwest National Laboratory. Using visualization techniques originally developed by the laboratory for the U.S. intelligence community, several disparate data types and sources are linked, transformed, and then presented on large computer graphic displays. These multidimensional spatial mappings portray information using complex visual patterns that humans can much more easily interpret than when the information is in the form of standard graphics, data tables, or text.
Known as the Visualization of Logistics Data project, the analysis of trends, patterns, and relationships in a large maintenance data warehouse enables logistics managers to capitalize on and exploit the ability of the human brain’s visual processing capabilities to rapidly perceive and absorb visual representations of large amounts of data in a manner not possible through listening or reading. This capability provides for a consistent and integrated picture of the health of the aircraft fleet, better parts forecasting, and reduced depot repair time and enables more informed decisions for work flow, scheduling, and resource forecasting (Lyons et al., 2011).
Finding 5-7. The potential for retrograde improvement using the Intelligent Collaborative Aging Aircraft Spare Parts Support project and the Visualization of Logistics Data project appears enormous. When used in conjunction with improved reverse logistics, these could pave the way toward a truly synchronized retrograde, enabling a responsive closed-loop supply chain with reduced requirement objectives and improved materiel availability and operational readiness.
Recommendation 5-6. The Army should adopt capabilities offered by both the Intelligent Collaborative Aging Aircraft Spare Parts Support project and the Visualization of Logistics Data project as first steps to
________________________
12 Nickolas Justice, Commanding General, U.S. Army Research, Development and Engineering Command, “Advancements in Soldier Power Presentation,” presentation to the Association of the United States Army Institute for Land Warfare Panel on Army Power and Energy Challenges on October 11, 2011.
incorporate predictive analytics toward a synchronized retrograde closed-loop supply chain. These concepts should be further extended, and adapted as appropriate, to sustain other fleets as well, including ground-based systems.
A shortfall in Army aviation depot-level reparable surge repair capacity in the existing Aviation Classification Repair Activity Depot organizational design and mission concept became evident in the early stages of Operation Iraqi Freedom. This recognition led to a new capabilities document for a Mobile Aircraft Sustainment Maintenance Capability (MASMC). This new concept emphasized both sea-based and land-based capacities to better and more responsively support an expeditionary capability, especially during the initial deployment phase, when seaports or airports may not exist or have yet to become available. There is historical precedent for this sea-based concept. The USNS Corpus Christi Bay was used as a floating aviation maintenance and repair platform off the coast of Vietnam from 1966 to 1975. The Marine Corps, traditionally an expeditionary force, has use of two aviation logistics support ships, owned and operated by the Navy, for aviation maintenance and repair support to the Marine Corps expeditionary forces.
A regionally aligned force structure could be adopted for the supporting multipurpose aviation sustainment brigades (MASMC, the replacement for Aviation Classification Repair Activity Depots). These brigades would be similarly organized but then tailored, trained, and deployed to conduct support operations in specific regions of the world. This concept of operation would associate each of five MASMC brigades with the five corresponding regional combatant commands: European Command (and the new Africa Command), Pacific Command, Southern Command, Central Command, and Northern Command. This arrangement would enable habitual association and command relationships to develop. It would increase the efficiency with which the MASMCs could meet their commands’ needs.
Finding 5-8. Resurrecting a sea-based maintenance and repair concept would be consistent with the Army’s evolution toward more robust sea-basing as a practical response to the growing anti-access, area denial environment.
Recommendation 5-7. The Army should re-establish a sea-based mobile repair capability for aviation and consider expanding the sea-basing concept to support maintenance and repair for ground systems as well.
Finding 5-9. Regionally aligned multipurpose aviation sustainment brigades would provide more efficient and responsive reverse logistics support to the major combatant commands.
Recommendation 5-8. The Army should adopt a regionally aligned force structure for multipurpose aviation sustainment brigades.
The concepts discussed above would form a resilient retrograde capability. They suggest the creation of prepositioned, mission-tailored support packages. The packages would be designed using readiness-based sparing and mission-based forecasting. If prepositioned packages are not used, the same effect could be achieved by setting aside small, similarly constructed packages that could be rapidly deployed along with the Army aviation unit; they would be similar to the Marine Corps flyaway element or the Air Force war reserve spares kit. Where existing (e.g., host nation) sustainment is not readily available, tailored mission support packages could be used to meet Class IX supply replacement needs at deployed locations. This would result in surplus inventory that could be used to meet any short-term demand surge that the existing logistics supply network infrastructure could not support (Parlier, 2010).
Further, in order to meet higher sustained demands during extended operations, resilient supply chain design principles would suggest either the creation of additional sustainment capacity or the moving of existing capacity closer to the point of need—or perhaps both. This concept would provide the ability to dynamically change the supply chain configuration in response to needs. Thus, the logistics network could respond quickly to a short-term demand spike with surplus inventory and then adapt to sustain increased longer term requirements by changing its configuration, relocating repair capacity closer to the point of need (Parlier, 2010).
Any effort to achieve resilience must focus on the strategic design and structuring of supply chains so they will be able to respond to the changing requirements of globally positioned forces constantly engaged in the conduct of a variety of operational missions in a wide range of environments. This ultimately requires innovation in supply chain design, implementation, and, especially, supply chain management. The topic of innovation is discussed thoroughly in the following chapter.
A significant proportion of materiel shipped to bases winds up as waste, and waste is generated on-site as well through normal living activities (e.g., by cafeterias, latrines, and showers). This waste must be disposed of on-site, typically by incineration, or shipped off-site. The challenges of waste disposal will vary depending on the size of the base; the energy content of the solid waste; and the presence of toxic wastes, especially dioxins, furans, and toxic metals in the waste streams. The disposal of military wastes by any method is also complicated by the potential for the presence of live ammunition in the waste streams. Waste disposal also depends on host nation laws and regulations. The current normal methods of waste disposal, as mentioned above, are incineration and removal from the base by truck. Incineration, however, creates a fuel demand to power the incinerator and increases the number of personnel at the base. Removal by truck, of course, creates a fuel demand for the trucks so used and ties up logistics resources that might be used elsewhere. It would be preferable to destroy waste on-site, recover energy from the process, and significantly reduce the amount of waste that needs to be trucked off-site.
Another obvious way to reduce waste is to generate less. The most straightforward to way accomplish this is to minimize the amount of packaging material used. Some items, such as munitions, have their packaging dictated by regulations for safety reasons. For other items, the volume of waste generated might be reduced by redesigning the packaging and packaging items more efficiently.
One concrete example presented to the committee was the various efforts to reduce meal-related waste. The committee learned during data-gathering that field rations—meals-ready-to-eat (MREs)—are commonly stripped for their protein and roughly half of the contents are simply thrown away. In response, the Army developed the First Strike Ration. This ration is nutritionally optimized and is designed to provide soldiers with meals that are compact, can be eaten on the move with no preparation, and used in high-intensity, mobile operations. One of these rations provides a full day’s subsistence, as opposed to three MREs, and it is supposed to eliminate the approximately 50 percent waste resulting from stripping MREs. Overall, the First Strike Ration provides a 49 percent reduction in weight and 55 percent reduction in volume over the MRE.13
________________________
13 Keith Bowman, Precision Airdrop Program Office, Air Force Research Laboratory, “AFRL Precision AirDrop (PAD) Update,” presentation to the committee on February 5, 2014.
Finding 5-10. It appears to be possible to reduce the waste burden on the logistics system by redesigning packaging, packaging items more efficiently, and minimizing any unwanted materiel so less waste is created in the first place, as demonstrated by the First Strike Ration.
There are many different approaches that can be taken to converting waste to energy. Depending on its size, solid waste may provide fuel to boilers that generate steam for electrical generators. Direct conversion of waste by pyrolysis or gasification can create solid, liquid, or gaseous fuels. Gasification or destructive distillation can provide gaseous fuel, such as methane, to power base diesel generators or even, under specific circumstances, vehicles or power equipment. Any resulting ash might also be useful. Bio char is being experimented with for water purification purposes.
The Natick Soldier Research, Development and Engineering Center has a program called Waste-to-Energy for Self Sustaining FOB/COP.14 This effort is to test a waste-to-energy system that will process 1-2 tons of mixed nonhazardous solid waste per day, producing environmentally benign products. The system is packaged in ISO/TRICON containers and has automated controls and operation, minimizing the manpower needed to operate it.15 The goal is to reduce or eliminate the need to remove waste from forward and contingency operating bases and to produce energy. The effect would be to reduce the amount of fuel that has to be delivered to bases for generators. There will be spiraling savings as less fuel means fewer trucks to deliver fuel, which means still less fuel.16
The Army’s Tactical Garbage-to-Energy Refinery, a technology development program, converts 2,000 lb into fuel to power a standard 60 kW diesel generator. Dry, solid wastes are pelletized, and the pellets are converted to syngas to power the generator. Sugar-rich wet wastes are processed using enzymes and fermentation. The result is hydrous ethanol (85 percent ethanol, 15 percent water). Mixing the ethanol and the syngas has resulted in the production of 55 kW using the 60 kW generator. It takes 6 hours to start the system, and during this time the generator uses about 1 gal of fuel per hour, about 5 percent of normal usage. The volume of waste processed by the system is reduced in volume by 30:1. It has been tested in Iraq (LaMonica, 2008).17
While the technologies to convert waste to energy in a civilian setting appear to be mature, a key challenge in using military waste to produce energy is the potential for live small arms ammunition to be present in various waste streams. For whatever reasons, such ammunition finds its way into waste, and becomes an obvious safety hazard to the personnel around such a waste conversion system. One committee member’s experience has been that this hazard has been an obstacle to testing and has necessitated time-consuming manual sorting of waste. This issue needs to be addressed before waste-to-energy technology can be widely deployed. Systems might be hardened to handle the cooking off of rounds, or the rounds can be sorted from the waste before it is used to generate energy. Also, soldiers could be trained to not throw their unexpended ordnance into waste receptacles or piles.
________________________
14 A FOB is a forward operating base; a COP is a combat outpost.
15 An ISO/TRICON container is a shipping container. When three of these are connected together they have the same footprint as a 20-foot ISO shipping container. ISO is the International Organization for Standardization, which sets standards for shipping containers, among many other things.
16 Richard J. Benney, and R.D. Carney, NSRDEC, and Edward J. Plichta, CERDEC, “Operational Energy—Advanced Woven PV, Equipment & Energy Technologies,” presentation to the committee on February 4, 2014.
17 The battle-tested TGER prototype has been improved since it was used in Iraq (Kristen Dalton, “Battle-tested TGER prototype improved since mission in Iraq,” Feature Stories, U.S. Army Edgewood Chemical Biological Center, December 4, 2012, http://www.ecbc.army.mil/news/2012/Battle-tested-TGER-prototype-improved-sincemission-in-Iraq.html).
Finding 5-11. Waste-to-energy technology holds promise for generating energy for forward and contingency operating bases. This technology will probably be less applicable to smaller bases and outposts. A key challenge to implementing such a technology is the presence of small-arms ammunition in the military waste streams.
Recommendation 5-9. The Army should act to eliminate the challenge of small arms ammunition in waste streams for waste-to-energy solutions. This could be done by developing hardened systems that can withstand ammunition cooking off, by developing efficient methods for the removal of ammunition from waste streams, or by training soldiers to not discard unexpended ordnance.
ARCIC (Army Capabilities Integration Center). 2010. Power and Energy Strategy White Paper. Fort Eustis, Va.: Army Capabilities Integration Center, Research, Development and Engineering Command.
AUSA (Association of the United States Army). 2004. The New Paradigm: Bringing U.S. Army Logistics into the 21st Century. https://www.ausa.org/publications/torchbearercampaign/tnsr/Documents/TB_SecReport2.pdf.
Choudhury, N. 2013. How green is 3D printing? http://www.rtcc.org/2013/09/02/how-green-is-3dprinting/.
DA (Department of the Army). 2007. U.S. Army CBM+ Roadmap. Washington, D.C.: Deputy Chief of Staff, G-4, Headquarters, Department of the Army.
de Neufville, R. 1990. Revisions of estimates. Pp. 279–285 in Applied Systems Analysis: Engineering Planning and Technology Management. New York, N.Y.: McGraw-Hill, Inc.
DoD (Department of Defense). 2008. Condition Based Maintenance Plus DoD Guidebook. Available online at http://www.acq.osd.mil/log/mpp/cbm+/CBM_DoD_Guidebook_May08.pdf. Last accessed September 30, 2014.
DoD. 2012. DoD Maintenance Fact Book: 2012. http://www.acq.osd.mil/log/mpp/factbooks/DoD_Maintenance_Fact_Book_2012.pdf.
G-4 Public Affairs. 2012. Army Launches Smart Operational Energy Use Campaign, Identifies 10 Initiatives. http://www.army.mil/article/89693/Army_launches_smart_Operational_Energy_use_campaign__identifies_10_initiatives/.
GAO (Government Accountability Office). 2003. GAO-03-705. Defense Inventory: The Army Needs A Plan to Overcome Critical Spare Parts Shortages. http://www.gpo.gov/fdsys/pkg/GAOREPORTS-GAO-03-705/pdf/GAOREPORTS-GAO-03-705.pdf.
Gavirneni, S., and T. Sridhar. 1999. Value of information sharing and comparison with delayed differentiation. Pp. 441-446 in Quantitative Models for Supply Chain Management, International Series in Operations Research & Management Science, Volume 17. S. Tayur, R. Ganeshan, and M. Magazine, eds. Norwell, MA: Kluwer Academic Publishers.
LaMonica, M. 2008. Trash-fed generator deployed in Iraq. http://www.cnet.com/news/trash-fedgenerator-deployed-in-iraq/.
Lyons, J.W., R. Chait, and J. Valdes. 2011. Assessing the Army Power and Energy Efforts for the Warfighter. Washington, D.C.: Center for Technology and National Security Policy, National Defense University.
McGee., V.E. 1971. Principles of statistics; traditional and Bayesian. New York, N.Y.: Appleton-Century-Crofts.
Nowlan, F.S., and H.F. Hemp. 1978. AD-A066579. Reliability-Centered Maintenance. Washington, D.C.: Office of Assistant Secretary of Defense, Manpower, Reserve Affairs and Logistics.
Optomec. 2006. Optomec Presents Final Report on LENS Phase II at 2006 CTMA Symposium in Williamsburg, Va. http://www.optomec.com/site/archived_news/news9.
Parlier, G.H. 2005. Transforming U.S. Army Logistics: A Strategic “Supply Chain” Approach for Inventory Management. http://www.ausa.org/SiteCollectionDocuments/ILW%20WebExclusivePubs/Land%20Warfare%20Papers/LWP_54.pdf?;%20U.S.%20GAO,%202003.
Parlier, G.H. 2010. Transforming Army Supply Chains: An Analytical Architecture for Enterprise Management. Pp. 69-93 in The Supply Chain in Manufacturing, Distribution, and Transportation Modeling, Optimization, and Applications, edited by Virginia M. Miori. Boca Raton, Fla.: Auerbach Publications/CRC Press.
Parlier, G.H. 2011. Transforming US Army Supply Chains: Strategies for Management Innovation. New York, N.Y.: Business Expert Press.
Parlier, G.H. 2013. Transforming a Complex, Global Organization: Operations Research and Management Innovation for the US Army’s Materiel Enterprise. Presentation to the 2nd International Conference on Operations Research and Enterprise Systems (ICORES) 2013, Barcelona, Spain, February 18. http://www.icores.org/Documents/Previous_Invited_Speakers/2013/ICORES2013_Parlier.pdf.
Pew Project on National Security, Energy and Climate. 2011. From Barracks to the Battlefield: Clean Energy Innovation and America’s Armed Forces. http://www.pewtrusts.org/en/research-andanalysis/reports/2011/09/21/from-barracks-to-the-battlefield-clean-energy-innovation-andamericas-armed-forces.
SAE (Society of Automotive Engineers). 2002. SAE JA1012. A Guide to the Reliability-Centered Maintenance (RCM) Standard. Washington, D.C.: SAE International.
SAE. 2011. AMS 4999A. Titanium Alloy Direct Deposited Products, 6Al - 4V, Annealed. Washington, D.C.: SAE International.
Sherbrooke, C.C. 1992. Optimal Inventory Modeling of Systems: Multi-echelon Techniques. New York, N.Y.: Springer Publishing.
Tuppen, S.J., P.E. Daum, M.J. Rawson, K.A. Lucas, and W.J. Brindley. 2006. Cost Effective Manufacture via Net-Shape Processing. Volume RTO-MP-AVT-139: 2-1and 2-26. RTO, Neuilly-sur-Seine, France.
Zhang, W., editor. 2010. Intelligent Energy Field Manufacturing. Boca Raton, FL.: CRC Press, Taylor & Francis Group.
