John P. Jackson, Jr. is a senior staff member at The Johns Hopkins University Applied Physics Laboratory (JHU/APL) and Project Manager for the Advanced Research Projects Agency (ARPA) Ship Systems Automation (SSA) program.
Bruce G. Coury is a senior scientist at JHU/APL. He received his Ph.D. degree in Industrial Engineering and Operations Research (IEOR) from the State University of New York at Buffalo in 1982 and his M.S. in JEOR in 1981. Dr. Coury has published more than 40 journal articles and conference proceedings papers in the areas of human factors engineering, human decision making, and user interface design.
If the Navy is to successfully perform the missions of the 21st century, affordability will remain a primary metric for deployment.
A substantial portion of the life cycle costs for naval combatants is associated with platform manning. An average of 30-50 percent of the direct annual operating costs for current naval combatants is budgeted for crew salaries and benefits. Examining the attack submarine force, assuming an SSN force level of 40 and a crew complement of 14 officers and 107 enlisted, this translates to $171 M being spent in direct operating costs (salaries and benefits) each year to man our SSN units. Additional direct annual costs of pipeline student billets, personnel administration, and logistics support drive the real price of manning even higher.
The significant manning reduction can be achieved by simply applying today’s technologies in a total ship engineering methodology. Several studies, including the NSWC Autonomic Ship study, have concluded that manning reductions of at least one-third are achievable on surface combatants by applying off-the-shelf technologies to a clean-slate, top-down design. Foreign navies have demonstrated that such manning reductions are realizable on submarines through construction and deployment of such platforms as the Dutch Walrus, German Type 212, and Soviet Alpha and Akula class submarines. The significant challenge to achieving significant manning reduction (up to one-third) lies in taking a total ship perspective in all design decisions.
While automation of operator watch station functions can improve warfighting capabilities. these efforts alone are unlikely to lead to significant platform manning reductions. Other duties such as administrative support, maintenance, and damage control may dictate that an individual sailor remain onboard even though his watch station workload has been drastically reduced or eliminated. Only by considering all the responsibilities of individual sailors, all watch organizations and department duties, and redistributing the workload amongst the remaining crew can automation lead to the removal of personnel from naval combatants. If new paradigms are purposefully considered along with emerging technologies within a total ship engineering methodology, even greater manning reductions might be realized.
If one were to ask the question, “What would it take to operate a submarine with 25 people,” new technologies, ship design concepts. policies and doctrine, and approaches to training would emerge. By establishing an ambitious goal and considering possible solutions with no interest or bias towards the status quo, new approaches and organizations can be generated. This approach was first introduced by Hammer and Champy as a methodology for revitalizing business organizations but is equally applicable to engineering naval warships. ARPA’s Maritime Systems Technology Office (MSTO) Ship Systems Automation (SSA) program has been applying this approach with some success in the development of new manning concepts for naval combatants.
Within the ARPA SSA program, new operating concepts have been developed for both submarines and surface combatants. Starting from a clean slate and considering all tasks performed by both operators and systems, one can identify critical tasks, mostly in the area of decision making, which operators must perform. Examining these critical operator tasks one can construct different crew structures that would support the execution of those tasks most effectively. The SSA program is taking these concepts further and developing prototype systems based on emerging automation technologies, then demonstrating the applicability of these systems in meeting the manning concepts. As the SSA program is successful it will be necessary to consider the other aspects of the translation (ship design, policies and doctrine, and training) within a particular ship class problem, and to evaluate these concepts with working systems at sea.
Particular functional areas on both submarines and surface combatants can be seen to drive the manning requirements for these platforms. Maintenance activities, damage control, and combat information processing constitute a substantial portion of the crew workload for both the DDG 51 and the SSN 688. Through the employment of condition-based maintenance with accurate monitoring and predictive models of ship systems, substantial reductions in onboard maintenance activities can be achieved. If accurate and precise assessments of catastrophic damage could be assembled and damage control personnel were given tools and technology that increased their brute force capabilities, reductions in damage control parties might also be achievable. And by improving sensor processing, information processing, correlation, and communication between information processing systems, ship workstation operators could be left to focus on decision-making vice communication, plotting, and filtering of data. While many other technologies are necessary to achieve a drastic manning reduction on Naval combatants, successful development engineering of automated sensor processing, information processing, intelligent systems interface, and brute force multipliers could yield significant cost savings on future existing combatants.
For several years ARP A and the Navy have invested heavily in the development of automated tactical sensor processing systems for automatic signal recognition, signal tracking, feature extraction, and signal classification. These algorithms and processing techniques have been applied successfully to many different types of signals, across a spectrum of frequencies including radar, ESM, and sonar. Several of these algorithms have been implemented in operational systems (including AEGIS and BSY-2), but have suffered from a Jack of computing resources and restrictions in system architecture. With the advent of massively parallel distributed processing architectures and high-performance array processors populated on standard backplanes, concepts for detecting and processing all detectable energy across all frequencies in all spatial bins can be considered. Prototypes of sonar processing systems that apply these advanced computing technologies have been developed and are undergoing evaluation within different Navy programs.
Intelligent sensors for internal ship monitoring are less mature than automated tactical sensor processing systems, but advances in MicroElectroMechanical Systems (MEMS) may make it practical to consider remote monitoring and control of internal ship systems. MEMS creates miniaturized versions of typical mechanical (flow, vibration, pressure) and chemical (temperature, constituents) detection sensors and collocates them with a microprocessor. Prototypes of these sensors have been developed and produced by ARPA on the same or similar assembly lines as standard microprocessors, suggesting the fabrication of such sensors could be inexpensive. As yet not demonstrated, future advances in MEMS could provide power-scavenging capability and wireless interrogation of the sensor. If both of these advances were realized, distributed, wireless intelligent sensing might be practical. As a proof of concept, the ARPA SSA program will be constructing, integrating, and testing in the next two years a distributed intelligent fire sensor for Naval combatants.
One of the primary tasks of operators working with today’s modern systems is the communication of data and information. Because many subsystems have been developed independent of one another, operators must act as intelligent links between subsystems, passing data between independent processing elements, maintaining associations between unlinked data, and ensuring consistency in the information representation across disparate elements of the system.
With advances in computer networking, inter-process data communication has been substantially increased, allowing larger amounts and different types of data to flow between processing elements connected to a common network. This allows one to consider peer-to-peer inter-process communications previously limited by functional priority and bandwidth. Expanding these concepts further, one can exploit the existence of a common network to consider central information access schemes and information managers between system elements. With effective information management tools, a collaboration between subsystems and operators can also be considered.
One of the key advances in the area of information management is the application of object-oriented design to inter-process communications and database management. By decomposing functions into generic classes and developing schemas for representing and organizing interface information, one defines not only a method for interfacing systems but a method for reasoning about a problem and a mechanism for interaction among disparate systems. Realizing that operators will remain the controlling element for all decisions, object-oriented approaches also serve to allow operators to interface and control multiple Advanced Reasoning Systems (ARS) simultaneously. A baseline implementation of an object-oriented information manager called the Central Information Processor (CIP) was developed by AT&T and demonstrated in 1994 as part of the Tactical Scene Operator/ Associate (TSO/ A) Prototype System demonstration, described later. Such systems do more than support immediate operating requirements but can also serve to adapt the system to support unanticipated requirements.
Information systems can be described by the flow of information between functional elements of the system. Traditionally, information flow diagrams are specified in detail and used to determine communication requirements between subsystems. Trade-offs between hardware processing capabilities and inter-process communication requirements result in an architecture of the system. But as the systems evolve and new techniques for data and information processing emerge, the interconnections between subsystems are modified, leading to costly system interface improvements.
By having an object-oriented information management approach as part of the system implementation, one can consider dynamic reconfiguration of information flows between connected subsystems, thus allowing new subsystems too. be introduced even when they directly impact the existing system functional partitioning and information flow. Through information managers, the decomposition of information might also be tailored to the capabilities of a particular operator, who in reality is an integral part of the system processing.
If one were to walk on to a modern-day SSN, high-performance computing systems and sophisticated processing algorithms can be seen in action. One would also see operators interfacing with these systems, supervising their operation, filling in gaps in processing capability, and integrating the results from multiple subsystems simultaneously. In this environment, operators not only perform manual manipulation of data, but monitor and control systems, interrogate the results of the subsystems, translate data into information necessary for decision making, and participate/communicate with other operators and supervisors in decision-making tasks.
In order to meet the goals of drastic manning reduction, it would be expected that many of the data manipulation, analysis and integration tasks performed by today’s operators would be embedded in hardware and software. The operators that remain would be responsible for monitoring and controlling systems, interrogating subsystems, and participating/communicating with other operators in making decisions. In effect, operators would supervise advanced reasoning systems (ARS) and automation.
In addition, with fewer personnel, effective workload/task management amongst operators becomes critical. Watch stations can no longer be dedicated to a particular task. Instead, operators must be considered as general resources, continuously engaged, shifting between roles as the situation changes. A new technology area of intelligent systems interface (ISI) is critical to successfully addressing these problems of supervising and interfacing with ARS’s and automation while fulfilling multiple roles in standing a watch or assigned to departmental duties.
In the coming months, ARPA will be initiating the development of technologies to support the functional requirements of the ISI. A significant challenge will be the incorporation of the intelligent systems interface (ISI) capabilities with existing advanced reasoning systems (ARS) and automation subsystems into working system prototypes.
Approximately one-third (132 out of 322) of the Condition 1 billet on a DOG 51 are assigned to damage control parties. Some areas are currently being automated including remote voice communications and some sensors. The significant challenge in a manning reduction for damage control parties and many special details is the reduction or replacement of the adaptable brute force capabilities of human operators. These tasks typically require the operator to directly interact with the task. Such tasks might include bulkhead shoring, maintenance and repair, and stowing stores during a vertical replenishment. While many of these tasks can be eliminated or made easier through modifications to the ship design, the development of technologies that would multiply the brute force capabilities of a small crew are necessary to achieve significant manning reduction goals.
There are many technologies being developed by the Navy, Army, and ARPA that might be applicable towards augmenting the brute force capabilities of the crew. The Army and ARPA under the Twenty-First Century Land Warrior (21 CLW) program are developing remote monitoring capabilities for soldiers on the battlefield to locate and assess the condition of personnel. Also under 21 CLW, unobtrusive head-mounted displays are being integrated for battlefield operations. These technologies might also be applicable to monitoring, assessing, and communicating with damage control parties dispatched throughout a ship.
Advances in robotic systems, including teleoperated mechanical arms and visual inspection systems show promise for removing the DC personnel from having to interact directly with fires and other hazardous conditions. Advances in lightweight protective clothing might allow firefighters to withstand prolonged exposure to high temperatures. While many of these technologies are under development, few have been evaluated for application in shipboard environments. In order to achieve significant manning reductions on naval combatants, these technologies and others must be demonstrated to support the damage control and special evolution requirements of Navy platforms.
There is a significant potential manning reduction on future Naval combatants. While technology development is occurring within ARPA and the Navy that could support the development of such a platform, established organizational structures and propensity towards the status quo limit our ability to consider large departures in ship design, policy and procedures, training, and ashore based infrastructure. New approaches are needed in these areas if a drastically reduced manned ship is to be pursued by the U.S. Navy. Such development is high risk and therefore possibly outside the bounds of the Navy’s current fiscal constraints. Through the development and evaluation of prototype systems and operational demonstrations of those systems, a proof of concept might be put forth that would provide a baseline for considering a drastic manning reduction in the near future.