In this modern world due to rapid development in technology there is an increase in the level of automation which makes human decision making critical in all fields. The goal of NCW is to enhance the ability of networks to provide shared information and situation awareness in the battle field between humans and automated agents. In this paper, we have discussed the interaction and collaboration of human with heuristic search algorithm in mission-missile planning and how the performance will show their impact in different levels of automation. Our aim is allowing human operators to manipulate the automated systems i.e. making the computational Algorithm working visible. This improves the efficiency and the success of whole mission . So an effective decision can be taken by collaborating humans and computers to discover a better solution than the one determined independently.
Keywords: Network-centric warfare, heuristic search algorithm , decision making , Toma Hawk Mission missile planning, Human supervisory control(HSC).
The change from the platform-centric (PC) to Network-centric warfare(NCW) has increased the level of automation and introduced new layers of human decision making complexity in the field of military operation concepts (Command and Control). NCW wants to hinge the enormous amount of information generated by NCW sensors which has to be filtered and synthesized to generate a critical decision under unknown situation with time constraints.
In the advanced level of automation in network centric warfare the human operators were unable to take optimal decision at critical condition under time constraints in a mission .So,there was a need to decrease the level of automation and give liberty to the human operators to implement new ideas in a battle field which may increase the probability of success because humans are natural optimizers where computer has some constraints in optimization .Taking the above factor as the background our paper has been designed.
Network-centric warfare's fundamental idea is to share information and collaborate humans and automated systems in order to enhance the quality of the information and shared situational awareness (DOD,2001).Information sharing in the NCW is critical factor because there are multiple sources which propagates information ,so there may be significant changes to the current situation .In this case there is a need of decision making under this particular constraint where the default algorithms used by automated systems are predefined which cannot produce optimized solution .Our aim is to use human in the decision making process and Automated system in the iterative part of the mission .In some cases human operators are also allowed to manipulate the automated systems where the efficiency and the success of whole mission can be improved.
Our relevant study in the network Centric-warfare leads to the argumentation in the department of missile matching .Missile matching in a Mission plays a very important role in NCW because the missile used in the mission may change significantly according to the situation in the battle field .If the automation level is increased in the missile matching department of NCW there will be rapid waste in the warheads due to default algorithms used for creating the automation .So there is a need to decrease the level of automation and use human operator for decision making so the whole mission optimization can be improved.
Information sharing and collaboration between both humans and automated systems in order to enhance the quality of information and shared situational awareness is a fundamental tenet of Network Centric Warfare (NCW) (DoD,2001).
As Network centric warfare is driven by advanced automated systems, so human-computer interactions are crucial aspect of overall performance of NCW. These interactions between human and computer are best examples of Human Supervisory Control (HSC).
HUMAN SUPERVISORY CONTROL (HSC):
Human Supervisory Control, is process in which a human operator has a overall control over the information displayed in the computer. When some operations are performed in the computer by operator the computer asks for information to the actuators in task environment . The information needed are collected by sensor in the task environment and sent to the computer and displayed to the human operator. This explains how HSC works by the process diagram below.
The following eight issues below are identified by Department of Defense(DoD) for Human supervisory control (HSC) relevant to network centric warfare(NCW) which can be considered as a significant barrier limiting NCW's potential benefits. (Cummings,Sheridan,Mitchell,2004)
- Appropriate level of automation
- Information Overload
- Adaptive Automation
- Distributed decision making through team coordination
- Complexity Measures
- Decision Biases
- Attention Allocation
- Supervisory monitoring of operators
As stated by the Deparment of Defense(DoD) in their 2001 Report to congress ,NCW is 'no less than the embodiment of an Information Age transformation '( DoD 2001). The intention of NCW is to change the magnitude of vision in the military operational concepts by effectively collaborating human operators with cognitive component such as unmanned air, ground and under water vehicles . Greater combat power is generated through the creation of shared situational awareness, increased speed of command, self synchronization and higher operational tempo, lethality and survivability (Alberts et al, 2000). The challenges of NCW is often misunderstood as increase in the level of automation is mandatory. In this paper, we are going to analyze how Missle Matching efficiency can be enhanced by decreasing the level of automation. According to the above criteria the Missile matching problem lies on the following concepts:
- Decision Making
- Levels OF Automation
Decision making in automated systems is always a challenging task because it deals with the cognitive capability .Developing such systems in NCW field is not an easy task .In Missile matching systems due to multiple information propagation by visualization tools, information over load may occur and lead to congestion.So there is a need of resynchronization process under critical time constraints .hence, NCW suggests human decision making in the case of situation awareness .Decision making process is explained in figure 3
Levels of Automation:
NCW plays a very important role in levels of automation because the increase in the automation may lead to less knowledge to the operators which may affect the missile matching as the information from sensors may change significantly . However, in systems like those that deal with decision-making in dynamic environments with many external and changing constraints, higher levels of automation are not advisable because of the risks and the complexity of both the system and the inability of the automated decision aid to be perfectly reliable (Wickens,1999).For strict tasks that require no flexibility in decision making and with a low probability of system failure, higher levels of automation often provide the best solution (Endsley & Kaber, 1999). Tasks requiring higher cognitive processes of planning and learning, like those needed in command and control, require knowledge-based behaviors and a critical element of decision support system design is recognizing which knowledge states would be lost if higher levels of automation were used.
The various levels of automation that can be incorporated into a decision support system are depicted in Table 1, and are known as the SV levels as they were originally proposed by Sheridan and Verplank (Sheridan & Verplank,1978).We are Using Level of automation2,3,4 in the field of NCW(Missile Matching ) for our analysis and found how decrease in the level of automation increases the efficiency of whole mission.
Problems with high and low levels of automation (Billings,1997):
HSC problems if LOA is too high
- Manual or mental skill degradation
- Loss of situational awareness due to lack of automation transparency, complexity, and inadequate feedback
- More advanced automation issues such as brittleness & literalism; in other words, the automated system might not be able to handle novel or unexpected events, as well as operate effectively in conditions near or at the edge of the intended operating envelope
- Time and difficulty to diagnose failures and manually take over
HSC problems if LOA is too low
- Cognitive and working memory overload in routine tasks under time pressure
- Human decision biases and heuristics
- Lack of repeatability and consistency
- Complacency and boredom
- Greater human interdependency and chaos when something fails, unless fail soft Safe guards are in place.
TOMA HAWK MISSILE MATCHING:
In this we are going to illustrate a mission missile problem which could be benefited by collaborating human and computer decision making .Our complete focus is on Tomahawk Mission-Missile assignment problem . Toma hawk is all weather subsonic cruise missile which is used for long range firing. This was introduced by General Dynamics in the 1970s and designed for low-altitude missile which could be launched from a submarine. A major improvement to the Tomahawk is the Network centric warfare capability
In Tomahawk we have three different types of warhead penetrating, unitary and submunition. These warheads are launched from US navy submarines for preplanned geographical locations which are called launch baskets. As these missions are pre determined before the launch date the actual missiles that are to be used are unknown until a hour before a strike because a significant changes consistently. Since there is a gap between mission planning and mission assignment there are lots of changes may occur at last minute ,as in old days mission missile assignments was done using a simple database called PC-MDS by the human operators which is used to display the mission database. As these assignments caused mission failure there is a need of automation .Personal computers were designed using which contains all missiles database such as terrain contour matching data ,global positioning system and DSMAC(Digital Scene Matching Area Correlation) PC-MDS is only a database of a mission are used by strike coordinators and assignments are done manually using either pencil or paper or their memory to keep track of the different factors and options to consider (Cushing, 2003).
From the above assignment problem a new interface was developed for strike coordinator to enter all the relevant information used in a mission such as missiles, targets, with some automated assistance .For each mission the assignment task was finding other alternative available resource (Missile) that matches the mission requirements.
Due to the complexity of the task result there is a need to simultaneously consider and equilibrate the three various information types which consists of hard constraint, Probabilistic information(Firing rate) and optimization requirements(Days to port) .For example ,two different missiles meet the same hard constraints ,in this case which is better to choose the missile that has firing rate of 85% and 25 days to port ,and the other missile 65% firing rate but 10 days to port. Under these constraints the level of automation plays a very important role because default algorithms cannot change significantly according to situation. Our analysis suggests LOA3 is feasible where humans are involved in the decision making due to the above constraints which can produce an optimized solution.
Interface1 (Figure 4) for the missile mission contains the database of the mission. In that database, mission table contains primary characteristics such as target, priority, route, launch basket ,Navigation equipment and missile requirements. Missile table characteristics are Missile name, ship name, launch basket, navigation equipment and warheads(unitary, submunition and penetrating) .At the bottom left of the interface 1 there is a table that shows the target warning and missile warning of target and missiles that are unusable and restricted. At the right side of the interface1 tool there is a horizontal four bar graph which shows the priority (high, low, medium) of the targets assigned according to the missile firing rate.
When operator clicks LOA2 button in the interface1, manual matching algorithm executes automatically between mission table and missile table .The resulting filtered missiles from the database are sorted in a new table popped below. When operator clicks on LOA4 button an algorithm generates instantly and the resulting sorted missiles are stored in the matching table. Heuristic search algorithm is used by LOA4 for sorting the missile by priority. As these algorithms uses high level of automation the human operator is running short of knowledge for missile matching assignment, was a huge problem.
Interface1 does not allow any collaboration between the human and computers in the important level of decision making. The interaction of human operator are used only in matching level and basic filtering level. As interface 2(Figure5) gives the human an over all control and the algorithm used here interact with the human. As such in the interface1 missile matching is done only using the manual matching algorithm which is done by the human operator in accordance with the criterions and ordered criterions which are listed next to the current assignment matching table in the figure5.
Auto match button which uses LOA4 is also available, but an addition feature is applied for LOA3 purpose which narrows the sorted missiles and interacts with the human in decision making .The hidden matching algorithm in this interface has been made to interact with the human by decreasing the level of automation was a major importance for the creation of this interface.
Additional feature to this interface is that it can react to significant changes according to the situation of the mission , since LOA3 is used. This interface has save option feature where current assignment is saved at the end of the screen and a new assignment can be generated according to the situation so that the comparison can be done between these assignment where the efficiency can be improved
Our analysis suggests that NCW's aim was not to increase the level of automation. Their idea or goal was to improve or enhance the efficiency of the mission by using computer support in the modern world..In this paper we have taken human interaction with computational algorithm as our main concept and analysed how decrease in the level of automation increases the efficiency of missile matching in the battle field. We have taken two interfaces of the battle field and analyzed how the drawbacks of Interface1 has been rectified in the interface 2.
Analysis of Interface1 & Interface2:
User evaluation was done by keeping two interface in display and generated a very critical feed back(Bruni & Cummings,2005).The main drawback of Interface 1 is, the use of LOA 4 for generating missile match. The computational algorithm used in interface 1 does not provide sufficient information to the human operator regarding how the mission missile match assignment is manipulated. So there is less knowledge to the human operator which is a huge draw back where human decision making plays a very important role in the case of situation awareness i.e. reacting to situations .The efficiency of the automation level was too less when used in the real time by toma hawk strike planners. The ability to react to situation in NCW which is a key factor in military operation was not satisfied .So there was wastage in warheads where missile matching was done by using automated algorithms which uses LOA4.
Interface 2 has the specific feature where the algorithm interacts with the human. The heuristic algorithm used in Interface2 provides collaborative work between human and computer .This collaborative work helps human in decision making, who will gain whole knowledge about the missile matching .Our analysis says this is possible only by decreasing the level of automation from level4 to level 3 which makes the human operator visible how the algorithm works or how the algorithm interacts with the human to produce the results which increases the efficiency of the missile matching. Another feature of this interface 2 is the facility of saving current missile matching assignment which is used in the case of situational changes in battle field .
In command and control, mission goals are driven by human intension and action which is then executed and communicated through automated technology i.e. computational algorithm. These algorithms with automation level 3 is used in our paper and discussed in detail with the interface tools. There are problems with low and high levels of automation .So level of automation must be in a way that is helpful for human operator in Missile Matching. So from the above analysis we propose that automated systems should be designed in such a way that every operation carried by the system must be visible to the human and also can be modified according to the situation which is a main factor in NCW.
- Bruni, S., & Cummings, M. L. (2005). Human Interaction with Mission Planning Search Algorithms. Paper presented at the ASNE Human Systems Integration Symposium, Arlington, VA.
- P.J. MITCHELL,M.L. CUMMINGS,T.B. SHERIDAN (2004)Human supervisory control issues in Network centric warfare
- DoD. (2001). Network Centric Warfare: Department of Defense Report to Congress. Washington DC: Office of the Secretary of Defense.
- Wickens, C. D. (1999). Automation in Air Traffic Control: The Human Performance Issues. In M. W. Scerbo & M.Mouloua (Eds.), Automation Technology and Human Performance: Current Research and Trends. Mahwah, NJ: Lawrence Erlbaum Associates.
- Cushing, J. (2003). A Measurement of the Effectiveness of an Overview Display for the Mission to Launch Platform Assignment Process Essential to the Tomahawk Strike Coordinator. Unpublished SM, University of Virginia, Charlottesville.
- Sheridan, T. B., & Verplank, W. (1978). Human and Computer Control of Undersea Teleoperators. Cambridge, MA: Man-Machine Systems Laboratory, Department of Mechanical Engineering, MIT.
- Endsley, M. R., & Kaber, D. B. (1999). Level of automation effects on performance, situation awareness and workload in a dynamic control task. Ergonomics, 42(3), 462-492.
- Billings, C. E. (1997). Aviation Automation: The Search for a Human-Centred Approach, Lawrence Erlbaum Associates, Mahwah,NJ.
- Madni, A. M., and Madni, C. C. "Context-Driven Collaboration During Mobile C2 Operations." The Society for Modeling and Simulation International 2004 Western Simulation Multiconference, San Diego, CA.
- Alberts, D. S., Garstka, J. J., and Stein, F. P. (2000). Network Centric Warfare: Developing and Leveraging Information Superiority, Command and Control Research Program (CCRP), Washington, DC.
- Sheridan, T. B. (1992). Telerobotics, Automation and Human Supervisory Control, The MIT Press, Cambridge, MA.