Engineering change management, risk management, manufacturing, ergonomics, information management, and modeling and simulation, all of which are addressed in this chapter, cut across the categories of knowledge, people, and property that are described in Chapter 1.
ENGINEERING CHANGE MANAGEMENT
In the development of a large vehicle system like Crusader, engineering changes occur at a rapid pace and at points up to and during actual manufacturing. To deal with such change, the automobile companies and manufacturers of earth-moving equipment, among others, have developed systems for managing extensive engineering changes. A change management system tracks engineering design changes to ensure that all members of the production team are provided with the most up-to-date design information for the system produced. Later, in support of the existing system, it will deal with technology insertion. Given the complexity of Crusader’s manufacture, UDLP will need to either develop or procure an engineering change management system for Crusader. Such a system should also address weapon system software changes, manufacturing process changes, and supply chain changes.
Recommendation 6: UDLP should pursue the development of change management techniques, particularly as they pertain to engineering changes, software changes, manufacturing process changes, and supply chain changes. The issue of technology change during the product’s life cycle should also be addressed.
When it comes to well-known commercial products such as automobiles or earth-moving machines, new products typically are based on an evolution of design, which entails much less risk than new manufacturing processes. There will be a lot of risks with Crusader—both in the various manufacturing processes and in ultimate product reliability—because every element of Crusader is new and untested. Through good engineering, these risks can be identified and minimized. It is not clear to the committee that UDLP has thought through the process of risk reduction.
Recommendation 7: UDLP should pursue risk identification, quantification, and management strategies as they pertain to the manufacturing processes. The manufacture of an all-new combat
system is fraught with risks that need to be identified and minimized before full production. Uncertainties are inherent given that Crusader is unlike any other system UDLP has produced.
Ergonomics is a technical area that is important to developing and maintaining safe and efficient manufacturing processes. Manufacturing ergonomics addresses such issues as safety at work, stress due to the environment and workload, workstation design, and the interaction of human operators with machines and tools. Manufacture of the Crusader artillery system poses its own ergonomic challenges. The challenges discussed here apply throughout the supply network, which will account for the roughly 80 percent of the manufacturing process that occurs outside the Oklahoma facility.
At the manufacturing level, the committee observed a positive environment with regard to safety and ergonomic issues through use of manufacturing simulations. However, room for improvement does exist, particularly with respect to training and human interaction with workstation tools.
Its experience in manufacturing has given UDLP ergonomic models for workstation operations, environment, and safety. However, it will need to identify strategies for human interaction with manufacturing tools, considering the sensory, physical, and cognitive aspects of such interactions. Most important, UDLP needs to use interaction protocols that enhance information input and retrieval, since the Crusader manufacturing environment is highly information-centric. The ergonomics of information management, such as efficient presentation of data to operators, is also important.
The committee also considered several operational issues related to human interaction with the workstation tools that were used for the Crusader project. The project should consider using commercial off-the-shelf interface technologies to support human operator use of information technology (IT) at the workplace. Human operators need to know the basic IT terminology for their individual workstations and across the system enterprise.
The methods for presenting information to the human operators should be such that the cognitive overload associated with information bandwidth, scale, display size, symbols, signs, and signals is minimized. The technology for information display and visualization to support workstation operations should take advantage of human sensory systems by giving the users display alternatives for both input-output and information processing.
The Crusader project should provide human operators with manuals and guidelines (in both hard copy and electronic form) to support the use of information technology. It should provide interfaces that enhance usability across diverse networks of people and machines.
The system interface should be scaled to support differences in workstation operation while maintaining seamless interoperability across the enterprise system. The use of adaptive interfaces is encouraged. A common system-level interface must be cognizant of the differences in terminology used by different members of the supply chain. For this reason, the means of communication and the interface models should follow industry practices, such as those of the Institute of Electrical and Electronics Engineers.1 In addition, such interfaces should be flexible and allow for tailoring to a particular user or organization.
As new information technologies are inserted into the manufacturing systems, human operators must be trained to use them. Policy guidelines are important for training the manufacturing
See the Institute of Electrical and Electronics Engineers Web site at <http://www.ieee.org>, accessed May 2002.
workforce to use the new technologies adopted for the Crusader system. Through simulation, UDLP has identified the ergonomic feasibilities of assembling Crusader, but beyond that it does not appear to have had the opportunity to deal with ergonomic issues.
The committee believes that UDLP should institute a continuous process of training human operators on the use of new technologies. UDLP should improve training in cognitive skills, such as problem-solving, since most IT work is cognitively driven. UDLP should also reorient and train human operators to measure their performance in information-centric operations.
Finding: UDLP has made an excellent start on dealing with the ergonomic issues associated with the feasibility of the final assembly process.
Because many entities are involved in the Crusader manufacturing operation, the organization is inherently complex, and the operation will require an effective communication and information management infrastructure. Based on the presentations, the committee believes that UDLP has embraced a strong course of action for implementing information management throughout the life cycle—for example, it has implemented programs to address manufacturing information needs at the conceptual level.
However, the committee noted additional opportunities for using IT to share information with the organizations involved across all the manufacturing phases. An earlier study identified a number of advances in commercial manufacturing—industry collaboration, adaptive enterprises, high-performance organizations, life-cycle perspectives, advanced manufacturing processing technology, environmentally compatible manufacturing, and shared information—that could greatly benefit defense manufacturing (NRC, 1999). The committee agrees that these advances would benefit the manufacture of Crusader and has also identified several information and communication technology needs that will be especially relevant to Crusader.
Electronic commerce (e-commerce) is essentially commerce in a Web-enabled environment. A virtual enterprise is a business that is built from organizationally and geographically distributed units but that for all intents and purposes operates as a single enterprise. E-commerce is an enabler of the virtual enterprise, because it facilitates cross-platform and long-distance communications. Given its elaborate system of contractors and subcontractors, the Crusader enterprise is a virtual enterprise. The committee suggests that the Crusader team implement software systems for collaboration and that it practice an e-commerce philosophy across all operation levels, including virtual and real information links between and for production workers, production processes, and the entire supply-chain management. The development of methods and tools for real-time information sharing, taking into consideration the colocation of stakeholders and data interchange standards, would prove helpful in this endeavor. Finally, the Crusader team may want to make use of existing virtual collaborative technologies through teleconferencing, telepresence, and other multimedia information systems, with particular emphasis on data interchange standards, information security, and manufacturing architecture elements when adopting Internet and Web services.
Recommendation 8: UDLP should aggressively adopt Internet and Web services to communicate information, share data, and conduct strategic discussions in manufacturing project management, thereby reducing the costs of travel for face-to-face meetings.
As noted previously, the success of large-scale manufacturing enterprise systems such as the Crusader depends in part on their ability to manage large-scale collections of data in heterogeneous forms. For this reason, common databases, which collect and provide data in standard forms, are
crucial. While UDLP has succeeded in generating and accumulating design and manufacturing data, these data are not available across the virtual enterprise. UDLP might find its mission easier were it to adopt a common data interchange standard and architecture for data exchange to support operations with its many vendors, both internally and externally. The committee suggests that UDLP evaluate current military, defense industry, and general industry standards for benchmarking information exchange. Finally, the committee suggests that UDLP define common variable names at the enterprise operating level for use across the Crusader manufacturing software system.
Another enabler of the virtual enterprise is intellectual property agreements that support the sharing of proprietary information across the entire enterprise. If this information cannot be shared freely, efficient operations will be seriously hampered.
Information security remains a very important component of national security. While the Crusader project is not classified as Secret, UDLP must recognize that design and manufacturing information could be pirated or information systems could be intentionally disrupted. Any system for managing information used in Crusader manufacturing should be designed and operated with close attention to information security. The committee suggests that UDLP use commercial off-the-shelf or in-house technologies to achieve information security and a seamless data environment for the Crusader network. The committee also suggests that UDLP implement a highly robust system for the security of enterprise-wide software to protect data, information, and knowledge of the Crusader system.
Recommendation 9: UDLP should develop intellectual property (e.g., trade secrets and competition-sensitive information) agreement models that protect information security while sharing it with suppliers in an extended enterprise.
At the current prototype level, UDLP has embarked on a vision of information management that is very broad and yet seems to be operating synchronously without an overall integration philosophy. The committee believes that architecture for managing manufacturing information should be adopted. By doing so, UDLP will position itself to manage a complex manufacturing system that is both highly adaptable and agile. This architecture should feature central management of a distributed infrastructure with integrated information and data models. It should facilitate coordination between organizations (departments, vendors, etc.). It should be network- and process-centric, with a highly transparent information flow to enable real-time system diagnosis and problem tracing. Finally, the architecture needs to be highly interoperable at and scalable to any manufacturing echelon.
The committee suggests that UDLP define and implement operational data capture technology for integration of machine and human sensory information, taking into consideration that information-driven manufacturing systems rely on automatic and multifunctional sensors and intelligent controls on the process and enterprise levels (NRC, 1998). UDLP should also define and implement a submodel within a larger information architecture for operations management (e.g., scheduling, transportation and shipping, or purchase orders) and define and implement a submodel architecture for configuration management that allows for real-time system reconfiguration (e.g., tool change, supply-vendor change, manufacturing location change, organization change, or policy change). Finally, the committee suggests that UDLP develop and implement a submodel architecture for logistics support across stakeholder organizations (including such operations as flexible supply chain management, characteristics models for each manufacturing operation, and improved design features).
MODELING AND SIMULATION
Modeling and simulation are important tools for the design and analysis of complex systems and can enhance understanding of system behavior by, among other things, identifying and resolving bottlenecks and operational constraints, as well as understanding operational- and system-level risks. This understanding also includes the cost/benefit analysis of component failures, part ordering policies, maintenance policies, and the effective use of scheduling policies. The committee observed that UDLP has extensive experience in implementing modeling and simulation in large enterprise systems and has embraced modeling and simulation to support the life-cycle management of the Crusader project. By looking at structural design, component layout, and human-system interfaces, UDLP has envisioned strategies to study the overall behavior of the system, including a common operating picture of the Crusader and the uncertainties associated with failures and maintenance.
It may be possible to obtain realistic production data from experience with similar projects, such as the Abrams tank and the Bradley fighting vehicle. The committee believes that it may be possible to validate models of the production system for UDLP by running simulation experiments in real, virtual, and/or constructive environments and comparing the results. Developing and running specific simulation scenarios to demonstrate the efficacy of hardware-software integration and interaction could also be useful in integrating simulation software and hardware.
The committee notes that there will be important challenges in shifting operations from Minnesota to Oklahoma just before the production phase of the program. With respect to this transition, UDLP should think carefully about what needs to be simulated and what kind of simulation is needed in each case (for example, visualization or animation versus stochastic behavior). Simulation can be used for several distinct purposes: validation and planning, monitoring and execution, or tuning and optimizing a supply chain. One tool is unlikely to be the best for all of these tasks. While high marks are given to UDLP for its modeling and simulation efforts to date, the committee notes areas where improvements could be made, including realism of simulation models, simulation software and hardware integration, human-computer interactions for the simulation environment, and validation and verification.
The current simulation models are developed on an ad hoc, stand-alone basis. In most cases, the predictions about model outputs or end states relevant to Crusader behaviors cannot be carried beyond the individual prototype applications. Integration of these applications may create new opportunities for their use. For example, it might be appropriate to determine the sensitivity of the manufacturing line-balancing model to the learning curves of the human operators.
A number of simulation models have been developed to test the behavior and manufacturability of the Crusader at the component levels. However, the simulation software and the hardware platforms have not been sufficiently integrated to achieve a common enterprise model.
Some excellent applications of modeling and simulation in developing manufacturing ergonomics were demonstrated to the committee. UDLP might consider testing the usability of simulations continuously throughout the Crusader life cycle, emphasizing usability knowledge across the cohort of users (designers, production personnel, and other stakeholders throughout the enterprise). Expert walkthroughs and usability tests often provide useful information for interface designers. Customizable interface designs that support modeling and simulation applications from different vendors but that are tailored and controlled by the system-level information management architecture may also be useful. The committee also believes that it would be useful to consider seamless multimodal human input technologies that could support real-time data input for modeling and simulation and improve system turnaround time, agility, and adaptivity. Such input technologies include the use of speech and the tactile and auditory senses (NRC, 1998).
If Crusader manufacturing is to exemplify an integrated enterprise system, as discussed in NRC (1998), models and simulations should include descriptions of the interactions between people and between people and machines. The integration of Crusader software models across all platforms
and user levels should consider the use of common metaphors and adaptive displays that capture the mental models of individual manufacturing stages as well as the use of seamless interface technology for Web and Internet applications.
Suitable practices for verifying and validating models and the resulting simulations will help to prevent problems when simulations are used in different environments. Human experts should be used to validate and verify manufacturing processes, including input data selection. UDLP should gather information on industry practices and compare it with the Crusader operational framework, and it should compare operational outputs with military and commercial standards in terms of performance variables such as cost, due dates, and system lethality, reliability, and ease of maintenance. The Crusader team should also consider the ease of integrating the systems for the business enterprise, manufacturing enterprise, and logistics enterprise. Modeling and simulation can be used to improve the effectiveness and flexibility of the operational networks used in manufacturing the Crusader artillery.
Simulation models developed in an ad hoc manner are likely to encounter validation and verification problems because they are sensitive to domain applications. The committee believes that such problems may occur in the existing subsystems model of the Crusader manufacturing system. These problems can be minimized, if not eliminated, if the validation and verification strategies have the following characteristics: common data models across the system context; common information scaffolding processes for simulation input-output analysis; a common modeling and simulation user’s manual; common strategies for modeling and simulation of information management; common standards for validating simulation; common performance metrics for the Crusader manufacturing environment; and specific application scenarios based on actual Crusader manufacturing information.
Recommendation 10: UDLP should develop an integrated enterprise simulation model to support real-time control of all levels of manufacturing, from the manufacturing cell or factory floor to the globally distributed extended enterprise. Simulation will ensure that operational issues are identified before they become problems.
The recommendations in this report should provide guidance to anyone hoping to participate in a large-scale manufacturing endeavor. More specifically, when followed, they should help ensure that UDLP’s manufacture of the Crusader systems for the Army is smooth and successful.