**2.5. Mission management system**

by the recovery of explicit feature-based descriptions of man-made objects [12]. Silhouette descriptions and classification are used to support the dynamic view positioning and auto‐

The algorithmic partitioning in a front-end digital signal processing (DSP) dedicates image acquisition, dynamic range equalization, image segmentation and region-of-interest identifi‐ cation [13]. The parallel processing engine supporting applications of a statistical evaluation of linearity or regular curvature have gradient extraction, edge thinning and threshold algorithms [14]. All the data from traditional navy data base, mine warfare environmental decision aid library (MEDAL) systems on MIW, MCM and tactical oceanography also could be accessed in main processing unit in the MCM UUV system and fed into the identification and classification processors [15]. A typical MCM data-processing flow of MCM UUV systems

Reacquisition and relocalization of predesignated mine or mine fields need huge amount of signal, communications and data packages from various sensor systems. Data sets for reconstructing three-dimensional (3D) and two-dimensional (2D) modeling are very big and very difficult to transfer through current acoustic carrier in underwater environments [5]. Some other information comes from distance oriented and directional angle of illumination of light source, which gives some incentive in the reconstruction of 3D or 2D model of mines.

Identification, reconfirmation of mines and mine-like objects (MLO) classifications are critical factors for mine disposal operations. Efficient and different bandwidth characteristics of communication careers are critically needed at the main control center of MCM operations to

matic object classification.

132 Autonomous Vehicle

is shown in **Figure 1**.

**Figure 1.** Mine warfare environmental data flow.

If the duration of MCM operations at sea is expanded to more than 50 days, it is necessary to maintain the clandestine nature of the MCM operation at enemy littoral zone; therefore, in these cases, mission management is critical to autonomous MCM operations. With the introduction of system autonomy of mission goals, which is a relatively new area of research, this system will retain clandestine operations and power system requirements for functionality [18]. Coordinated MCM mission management systems optimize available sensors and systems, regardless of the host platform, to ensure that the most effective is used when and where it is most needed.

Fundamentals to the MCM operational concept are to locate minefields, identify no-mine areas accurately and clear mines efficiently as soon as possible [19]. This area focuses primarily on unmanned autonomous vehicles intelligence since these often have the greatest redundancy, and because they have the most intricate machine-readable, autonomous mission plans. Models of the vehicle including their planning/control systems, and operating environment can be linked together to form an assessment tool [16]. This tool helps to verify interface compatibility and basic functionalities, as well as to facilitate exploration of control options and evaluation of potential benefits.

The mission control data, which are required to define current mission objective, the vehicle's dynamic characteristics and the environmental data are collected from external sensors and provided by the user, as they are specific to the effective MCM operations. The autonomous mission-planning algorithm translates the mission requirements into a mission plan, a set of constraints and execution sequences. An integrated mission planner and power management algorithm would combine to this intelligent system with motion and power control subsystem [10].

MCM mission management configuration [17] of MCM UUV consists of degrees of perception, intelligent control and management to define sense-plan-act or sense-react behavior doctrine. Functional limitations of vehicle sensor systems imposed by the combat environment require alternative course of vehicle control, which defined mission goals to be factored in the control system appropriately. A common approach to mission implementation process is to develop functions, which are grouped and sequenced to perform overall tasks through artificial intelligent (AI) expert system architecture [12].

It is based on a traditional perception-action loop, allows low-level sensor signal processing and does feature extraction to assist mine classification, MCM mission planner and vehicle control system. MCM UUV mission management system focuses on mine classification and dynamic repositioning, which optimizes the aspect relative to the designated mine target and clearing mine. Obstacle avoidance navigations require relatively sparse sensor data in the dynamic construction of world models for environment [12]. In the main control block, the perception/classification processor is combining with dynamical position sensors via a highly maneuverable platform. MCM expert system has the capability to perceive the object of interest from multiple perspectives, which then increases the probability of classification for the mine targets. A key concept in data fusions of expert system includes the employment of heuristic knowledge, which incorporates constraint knowledge associated with target characteristics, environment and attitude of vehicle.

This provides basis for interaction with the object of interest, and dynamic perceptions that provide active sensor management and vehicle path planning in the navigation and guidance [16]. A second unique aspect of the AI expert system architecture is the implementation of sensing, evaluation and action encapsulated as subordinate tasks under the integrated mission control system. It optimizes machine-generated direction of task-organized approach and initiates signal for the course of action [6]. The MCM AI expert system architecture and its relationships between the executors, and the data/signal processing are shown in **Figure 2**.

**Figure 2.** Operations of MCM mission management expert system.

#### **2.6. Vehicle management system**

The system capability of precise navigation and operational data collection is critical to ensure safe navigation of the vehicle and in the achievement of system objectives. To resolve the vehicle position at the submeter level, a compact low-power solid state inertial measurement unit (IMU) has been incorporated [20]. This unit measures the change of 3D velocity, angular velocity and temperature as well as performs corrections of thermally induced drift. The IMU is augmented by a compact Doppler Sonar System using the Kalman filter or other processing techniques. A precise navigational compass and clinometers provide heading information and navigation frame correction data for noises [21].

Simultaneous localization and mapping (SLAM) techniques [22] are utilized as a navigational tool and are adapted for reconfirmation of designated mine localizations with autonomous navigation for obstacle avoidance to a safe margin. All the information are filtered and processed in a data processing and distribution unit and distributed for navigation, SLAM processing and mine neutralization procedures. With updated environmental 3D map and obstacle information, the MCM UUV navigation system can be guided and controlled within guided optimal paths to the targets with a degree of accuracy [22].
