One great aspect of being in the Submarine Force is that we ship-drivers have always been firmly supported by the very latest, most advanced technology that the scientific and engineering communities can provide. One visit to any of the technical symposia, research labs, or submarine industrial partners is immediately convincing. The powerful mixture of talent and passion is palpable – you can feel it everywhere. It’s no wonder that the relationship between the Undersea Enterprise and our partners in technology and industry is the envy of the defense department.
Focus on the Commanding Officer’s Decisions
Advances in technology have enabled clear improvements in capability and performance across all aspects of submarining. In particular, the rate of technology insertion into the submarine’s tactical systems is currently faster than ever before- optimally tuned to Moore’s Law and production life cycles. My experience as Commodore, Prospective Commanding Officer Instructor (PCOI}, Deputy Squadron Commander, and Commanding Officer have convinced me that if we’re to translate improvements in technological capability into improvements in tactical performance, we must focus on technology as a tool to improve command decision-making. At the end of the day, it’s the Captain who will make the key decisions as to how to employ the ship. Thus, when introducing a new tool or system, the question must be asked:
What command decisions are involved in the scenario, and how is this system going to enhance that decision-making and improve performance?
Anyone who has been at sea can attest that capability and performance are related, but are quite different things.
A quick discussion of decision-making will help formulate a systematic approach to answer the above question. For the purposes of this discussion, we will roughly define two general types of decisions: analytical decisions and intuitive decisions.
To make analytical decisions, one weighs several options, and decides on the option with the best balance of risk and gain. This type of decision-making is well understood, and is used often by submarine COs. While this is a necessary strength for Command, it is not sufficient, and is not a good predictor of tactical performance. In fact it is not uncommon that very good analytic decision-makers struggle in the attack center, where more complex intuitive decision-making becomes important.
Intuitive decisions are made after one detects the cues and patterns that emerge from complex situations, and then chooses a single course of action that will likely be successful. The action chosen is based on experience -the person has seen similar situations, and has a “library” of pre-planned responses (mental models) from which to draw. Based on recognizing the situation that faces him, the decider quickly converges on a single course of action, and runs a mental simulation of the action. If the simulation ends with success – he executes that option. If the simulation is not successful, he quickly makes adjustments to correct the difficulty or tries another model altogether, running through the process again, until he finds a successful course of action to take. It’s important to realize that intuitive decisions are made quickly compared to analytical decisions, and that the decider is not comparing options. If the first projected course of action works – he executes.
Knowledge of this type of decision-making is not so well understood, but has applications in most tactical and maritime scenarios. As a simple example, a CO may recognize the patterns emerging from a crossing situation (“That contact has a zero bearing rate and port angle on the bow, and will collide with me if nothing is done”). He then projects a mental simulation of his action based on the “mental models” he has developed through his experience (“I should turn to starboard now”). If the projection results in a satisfactory result (“I will get off his track by 2000 yds, and he will pass safely down my port side”), he executes his decision. If the projection does not have a happy ending (“I will run aground on that beach to starboard”), he chooses another option to consider (“I should slow by backing down and let the contact pass ahead.”). Even in this simple example, one can see that there are several correct courses of action. The CO, by virtue of his experience, can quickly converge on a course of action that will work. I have borrowed this model for intuitive decisions from Dr. Gary Klein’, which serves to provide a useful structure in enhancing intuitive decision-making. This model will serve as a reference for the rest of the discussion.
There are a number of procedural structures already in place to make command decision-making as robust as possible. As taught in the Submarine Command Course and on the waterfront, the tactical decision (or execution) phase is supported and enhanced by a cycle that begins with detailed analytical planning. The plan is then communicated to the operational team by a comprehensive briefing scheme. Then, once the operation is complete, the decisions are examined in detail by a rigorous assessment. The conclusions from that assessment are then fed back into the process, to enable the decision-maker to do better the next time he faces that situation.
Throughout, the decision model is used to focus each phase. The planning and briefing phases are focused on anticipating the expected cues and patterns that a situation may offer. A set of pre-planned responses (courses of action) may be sketched out. Experience is key, and all members of the command team are expected to leverage their experience in the preparatory phases so as to support optimal decision-making when the operation begins. After the evolution is complete, the model again serves as the focal point of the assessment- what patterns and cues were missed? Where did our library of pre-planned responses fall short? And so on.
Acting like the force of gravity on all of the command’s efforts is a risk-management priority structure that mandates that one must first ensure safety, then tactical security, then once these are finely established, strive to achieve the aim of the mission. This “submarine risk management pyramid” also serves to maintain focus in the decision-making phase. Should a CO, during the execution of his mission, find that he is at risk of being counter detected, he must first guarantee stealth – even if this means suspending the mission until he is once again secure. If things degrade further and he finds himself in jeopardy, then he must establish safe submarine operations as a top priority -even at the risk of compromising stealth. Thus as operations unfold, a submarine CO can find himself dynamically moving up and down the pyramid in response to various circumstances.
The figure below sums up the decision-making discussion as it is applied to submarine operations. Submarine operations are complex enough to require a combination of analytic and intuitive methods. Any team that does not do their analytic “homework” will be ill prepared for the complex decisions that they will face during execution. In general, we prepare our watch officers initially to be safe -supported by a thorough knowledge of submarining knowledge and procedures, an analytic process. For their first watches, junior watch officer’s decision-making is largely based on the analytic work officers did to qualify. No wonder they are monitored closely during their early watches! As they become more experienced, they are able to recognize and manage more complicated scenarios – recognizing more subtle patterns, and building a bigger library of mental models. For instance, they will become adept at maintaining not only safety, but also stealth. As the situation becomes even more complex, experience plays a correspondingly larger role. Eventually, at the most complicated levels, the Commanding Officer may be the only one on board with sufficient experience. He will be guiding the operation – either personally or through his more experienced watch officers – deftly leading his team up and down the submarine pyramid as required to guarantee safety, security, and mission accomplishment.
Starting with the decision model; considering the planning, briefing, and assessment structures that support the Commanding Officer’s decisions; all of which are influenced by the hierarchical priorities of safety, stealth and mission; one can begin to see several roles for technology.
For instance, technology must be used to present the situation to the decision-maker. Taking this presentation a step further, it is possible to enhance the important patterns that characterize a situation (and to suppress the noise that can hide the important cues and patterns). Similarly, there is a valuable role in predictive simulation -allowing a team to try out or practice different courses of action in a simulated environment before entering the actual tactical situation. Finally, in the heat of the moment, once a CO commits to a course of action, technology can expedite that course of action in a streamlined manner-minimizing the required effort to translate the idea into action.
In reality, it is much more dynamic than that. The description above generally relates a scenario where technology comes to bear in discrete points along the decision cycle. The truth of the matter is that processing and display power have progressed to the point where there is a constant ‘dialogue’ between the decision-maker and the supporting technology at all points of the process- and in all phases of planning, briefing, deciding, and assessing. Instead of a “human-machine interface,” it is now more accurately a conversation between partners-the decider and the system. An image that might help is an early scene from the movie “Minority Report,” where Tom Cruise, playing the role of a law-enforcement officer, interfaces with a futuristic computer to sift through and mine a complex and diverse data set to determine the details of a crime that has been predicted. His goal is to intervene to prevent that crime. The situation facing the submarine Commanding Officer is much the same: he also must sift through lots of data to make sense of what is going on around his ship, and then to act to shape the future of that situation in the manner he desires. Like all good science fiction, the dialogue between Tom Cruise and his future-computer is both entertaining and educational. We are not far from realizing that level of interaction today.
Some Qualities of a Good Partner:
Confidence and Communication
A trusted mentor often tells me “submarines are like old English villages – predisposed to be suspicious of outsiders.” This is true! The combination of a hostile operating environment and the intolerable price of failure has produced a community with extremely high standards. We need to be convinced of our “partner’s” utility before we begin to trust them. To continue this thread, it may be useful to discuss the attributes of technology in terms of those qualities that a Commanding Officer might desire to have in a partner that is focused on solving a common problem. Many of the qualities that set apart good partners from bad fall broadly into two categories: Confidence and Communication. If the decision-maker can be co11fide11t that the technology will perform as desired when required; and if the can quickly and easily communicate ideas to the machines, and just as quickly and easily understand the results of the machine’s contribution; then there is the foundation of a healthy dialogue. If there are problems in either of these areas, the relation-ship is in trouble.
Confidence Begins with Transparency
Systems placed on submarines must be fully checked out-examined from all perspectives. It is a great strength within the Undersea Enterprise that we are not afraid of a thorough inspection -in fact we welcome it, and the lessons we learn from it. This must be true where technology is concerned, and it must include transparency in all aspects of the process-development, engineering, and implementation. Submariners detest black boxes. It is insufficient that the system comes with an externally stamped seal of approval when it arrives. The algorithms, sensors, displays, materials, construction and testing that characterize our systems must be thoroughly evaluated and understood not only by experts but also by the operators. After all, it is the on-board decision-maker-the Captain-who is the person who must understand the system well enough to trust it underway in a tactical situation. It may be certified for use by external engineers, but it will only be actually used when it gains the trust of the Captain and his team.
Related to transparency is need to know the principles and intermediate steps behind the final answer. Submariners embody the adage that “the devil is in the details.” We do not accept the ‘Q.E.D.’ unless we know the steps involved in the proof. Systems that only serve up the final answer can eventually make us stupid-and even unsafe. Relying on the final answer will lead to atrophy of our healthy tendency to question initial conditions and assumptions, and we may lose our ability to know when the answer makes sense or not. I have seen examples of this in tracking exercises. Evaluators managing digital plots, where the computer does the line-of-sight mathematics behind the scenes and displays the final result, will fail to recognize or question when the plot clearly does not make sense -when it does not match the system solution. To address this phenomenon, for no system is infallible, we at DEVON 12 are working on producing a set of primers, designed for the junior officer, that would focus on the principles and fundamentals behind processes like TMA, Search, and the Sonar Equation. Our hope is that these primers will complement the great work of the Submarine Learning Center and on-board training programs to maintain this critical knowledge of fundamental principles.
This requirement to know the steps leading to the final answer also has important implications for Tactical Decision Aids (TD As). Using vast computing power, these spectacular tools can evaluate complex situations (sound propagation in shallow water), and make valuable recommendations about the best sonar lineup to use to achieve the desired aim. To image that target on sonar, what frequency range is best -HF or MF? Should it be active or passive? And, once these big questions are answered, what are the specific settings required to optimize performance? Computers can greatly assist in answering these difficult questions-as long as we keep in mind that in the end, the CO must make the decision-because only he can properly assess risk to his mission and crew. So, TDAs should enable an operator or supervisor to walk the CO through the process that led to the recommended mode. This walk would allow the CO to check assumptions, ask the right questions, and in the end have confidence that the recommendation is sound. Only then will he use it.
There’s Gold in the Raw Data
Another role of TDAs is to reduce complex data to enhance operator understanding-to boil things down to the essential elements of the problem. There is a vulnerability here: critical information is often in the raw data–information that will be lost by over-smoothing. Just think of the difference between the sonar broadband display (AVSDU) and the Contact Evaluation Plot (CEP) derived from that display. While the CEP displays a lot of important information for decision-making, there is a lot MORE information on the A VSDU. In addition to the bearing-time history of contacts, there are all the trace-dynamics of each contact-which ones are the brightest (loudest), which ones are just whispery traces, which ones are transients, etc. To rely on just the CEP would be to lose all this additional data- which could be critical to making a proper decision. So while the CEP and similar TDAs have an important place in decision-cycle- to enhance certain patterns – the raw data may often contain the essential pattern key to deciding what course of action to take. It’s important to reduce complicated data to a set of essentials. But it’s equally important to retain the raw data, for we often don’t know what bits of information will prove to be the deciding factor in a given scenario. Boiling down data generally helps the junior decision-maker to become more effective sooner. Retaining the raw data allows the experienced, expert, decision-maker to see the subtle patterns that only he can see (it is no accident that Submarine COs make most of the trace dynamic classifications of submarines while looking at the A VSDU), and to explore his hunches. The emerging trend in tactical systems is to do both-provide data reduction and smoothing, then superimpose that smoothed data on top of the raw data. In this way, we enable both beginners and experts.
It’s imperative that these systems be designed to support command-level decisions. We must fight the tendency to design displays and interfaces that are solely optimized for the operator. While it’s important that the operator can use his display, at some level the system must serve the CO, otherwise it may prevent the most important decisions from being made. A quick example will illustrate what I mean. For discussion purposes, a TOA that assesses the ship’s non-acoustic vulnerability, in real time, for a given mast configuration, would be a valuable tool. As discussed before, stealth is a fundamental step in the submarine pyramid. If we place this TOA in radio, it will no doubt provide the operator with a much richer understanding of the electromagnetic environment and the ship’s vulnerability. But the operator is not the one who will make the key decision-to raise or lower all masts! That decision is made in Control, often at the command level. So the TOA properly belongs in Control, in a format that can alert the CO (who is not sitting in front of the panel as a dedicated operator), that he’s got a problem.
Technology that serves the decision-maker is fully transparent at all levels, and is designed with the recognition that the decision-maker should have access to the fundamental principles and raw data behind the final answer. It serves decision-makers because it performs in a way that garners their trost, consistently arriving at the answer in a way that inspires confidence.
The value of any partnership is the power that arises from the communication between the two partners. If, as in our case, the partnership of the CO and his technology is focused on solving a problem, then at the most basic level, the conversation must keep pace with the problem at hand. It does no good to solve for the contact’s solution after he has faded and opened up outside of weapons range. The channels of communication between the two must enable the rapid cognition required to keep pace with the problem as it unfolds.
Captain to the CONN
It’s 0230, and the tactical situation has become too complicated for the Officer of the Deck- he is overwhelmed. He knows he needs help, and buzzes the Captain out of the rack. The CO gets up out of a dead sleep, and walks into Control -he knows it’s probably not going to be a good scene. Yet even in this complicated scenario, the CO will enter Control and first look for a few key parameters to grasp an initial sense of the situation -and these parameters are fairly common for all CO’s. I’m going to refer to this set of key parameters as the “vital signs” of the problem. The analogy fits. The human body is also an incredibly complex system, with thousands of processes underway at any one time. Yet, when a doctor approaches a patient for the first time – even in a trauma scenario (like our “Captain to the Conn” scenario)- the doctor will want to know the vitals: pulse, blood pressure, respiration, etc. Anybody in medicine knows the drill.
What are these tactical vital signs? For this discussion, it’s not important that they’re conclusively defined- every reader will have their own set of them, and they’ll be slightly different depending on his situation. For this discussion, I’ll assume that most sets of vital signs would include own-ship’s course, speed, and depth. These parameters, because they are the CO’s vital signs, should be consistently displayed in a way that visually prioritizes them- makes them easily and instantly available to the CO. We should strive to display them in a common field, in a common place, in consistent format, on all displays. So that, in the heat of the moment, when the CO is trying to quickly make sense of a complex situation, he will not have to think about where the vital signs are displayed-they will leap out at him! This seems fairly obvious until one considers even our simple vital signs: own-ship course, speed, and depth. On our current systems, depending on where you look -sonar, fire control, ship control, etc., this data is displayed in different fields, in different places, and in different formats. As PC01, I often found myself hunting around for the data I needed, until I had “calibrated” myself to the system and display that I was using. I often wished that the vital signs display had been standardized. We are just starting a discussion about what data would constitute the vital signs in a given situation, and how this data should be presented to the decision-maker-the Commanding Officer-to optimize performance.
Don’t Stress Me Out
In the PCOI scenario above I always found the vital signs, but it was frustrating, and more important, it was stress-inducing. This leads to the next attribute. The dialogue between technology and the Commanding Officer is never more important than at times of high-stress. In a real tactical scenario, when the submarine may be in-extremis, or at the firing point between two submarines at near-parity, there is a good chance that the scene in Control will be rapidly changing and high-stress. In this scenario, decisions will be made rapidly, and will need to be executed just as rapidly. This is the realm of intuitive decision-making. Experience counts, as it enables one to see critical patterns quickly, have a ready-made course of action, and execute. This is the litmus test. In this case, technology should strive to minimize the stress involved. At the very least, it should add no stress! The vital signs of the scenario should be easily and intuitively displayed, and the resultant course of action should require minimal effort to execute.
In the Submarine Command Course, we aim to put the students under stress to see how they respond. And while the primary aim of that exercise is to teach the student how to better handle these situations, I have seen both Commanding Officers and operators wrestle with their systems as they struggled to gain control of the situation. Critical patterns were often difficult to see in the confusion, and many button presses, keystrokes, screen-touches, mouse-clicks, and other manipulations were often required to execute the desired course of action. We’re on a good track to make this situation better, and we need to continue to make progress.
Machines Can’t Make Us Experts
No matter how intuitively a particular pattern may be displayed, and no matter how streamlined the desired course of action may become, in the end it’s operator experience that matters. Without experience, the CO will not know what he is seeing, and will never gain a sense for the complexity of the situation. Technology will never be able to replace experience. But it can help build experience, and the current systems are better than ever in that regard. New processing and display technology enables embedded training modes and simulation that were unapproachable until just recently. With a very high degree of fidelity, it’s possible to simulate just about any scenario that one may want to see. The screens and displays will provide nearly the exact response that a team will see at sea. By cleverly designing the training scenario, it’s even possible to create the stress that one might feel during the real thing! The best training uses these trainers in scenarios that are chock-full of opportunities to detect cues, recognize tactical patterns, then formulate and try out different solutions. Once done, the new trainers can allow the team to replay the scenario to assess their performance. They can see the patterns that they missed the first time through, and explore alternate course of action to build up their library of mental models. So while machines will never replace an experienced decision-maker, technology is an essential tool to training in decision-rich scenarios that build experience and intuition as efficiently and effectively as possible.
USS VIRGINIA – A Decision-Centric Ship
There can be no better way to conclude this paper than to discuss the latest class of warship to join the fleet. To see a vision of the power of decision-centric technology, one only needs visit the USS VIRGINIA (SSN 774). I am privileged to have that submarine in SUBDEVRON 12, and have therefore had many chances to see the ship and crew operate together. In fact, it was while watching the command team on USS VIRGINIA that the image of a continuous dialogue first came to mind, for that is exactly what happens on that ship. The entire ship is designed to provide the command with easily understood information, and then to enable the resultant course of action with almost no unnecessary effort. Much has been written about VIRGINIA, and much more will be written as we learn more about her. Suffice it to say here that she already embodies many of the principles discussed above. It is extraordinary how fast information moves around that ship and gets to the person who needs to use it. Enabled by the superb design and engineering inherent in the ship, the Commanding Officer and his crew have reduced processes that are fairly complex on a Los Angeles-class submarine to much simpler and more streamlined procedures, often handled by much fewer people, with noticeably less entropy. The training modes embedded in her Attack Center enabled the crew to do most of her tactical training not at the schoolhouse, but on board -using the exact equipment she will use underway. Ship control, using fly-by-wire techniques, provide just the right balance of assistance while permitting operator control; the displays are intuitive and very quickly understood. This decision-centricity has enabled her crew to train up and deploy years ahead of the initial schedule. She is on deployment as I write this, making decisions that are translating technological capability into tactical performance fighting the war against terrorism. That’s a mental model worthy enough for all our libraries.
Perot Systems Corporation is an internationally recognized provider of technology-based business solutions. We employ more than 16,000 associates and in 2004 generated revenues of $1.8 billion. Through innovation and a focus on customer confidence, our associates drive quality results for our federal clients. Services that we provide to the Submarine Force include:
• Program Management
• Acquisition Management
• Financial Management
• Logistics Management
• Quality Assurance
• Maintenance and Engineering Support
• Submarine Safety Support
• IT Solutions