PDF Download 
N. Vedachalam, “Cognitive Digital Twins in Strategic Anti-Submarine Warfare: A Scoping Review,” ORF Special Report No. 268, July 2025, Observer Research Foundation.
Submarines are central to naval campaigns and, with growing cruise and ballistic missile capabilities, play an important role in deterrence, subsea warfare, and land-attack operations. As the most survivable leg of the nuclear triad, sea-based nuclear deterrents lead global powers to deploy the majority of their nuclear warheads on ballistic missile submarines, ensuring a reliable and secure second-strike capability.[a] Even before conflict arises, a submarine’s ability to project power undetected offers strategic advantage. Tactical submarines also provide stand-off capabilities for conventional deep-strike operations that could be used in a land-based attack. While the speed, endurance, quietness, and stealth features of submarines are on the uptrend, the development of Anti-Submarine Warfare (ASW) weapons and systems with the abilities to neutralise an opponent’s underwater force is a sine qua non.[1]
Early, first-generation ASW tactics relied heavily on static defences, such as underwater mines and chain-link nets. Second-generation tactics evolved with the invention of the hydrophone. Post-1980, third-generation ASW systems emerged, including Surveillance Towed Array Sensor Systems[b] (SURTASS), where ships trailed long hydrophones capable of transmitting acoustic intelligence via satellite to ground stations.[2] Present-generation ASW systems feature Intelligence, Surveillance, and Reconnaissance (ISR) capabilities supported by an interconnected network of surface ships, aircraft, and submarines.[3] Strategic uncrewed ASW is expected to move from platform-centric control to autonomous systems with collaborative control and cognitive capabilities, enabling persistent maritime surveillance, intelligence gathering on submarine movements, countering asymmetric threats, and neutralising hostile[c]submarines.[4]
While reliable detection, tracking, and intelligence gathering on hostile submarines remain central to undersea warfare, six major challenges hinder ASW operations. These include the proliferation of submarine operators—42 countries now have submarines, driven by security objectives, threat perceptions, regional dynamics, and strategic relations between global powers. Submarine counts for North Korea, the United States, China, Russia, Japan, South Korea, and Iran are 71, 67, 59, 49, 22, 19, and 17, respectively. Other challenges include increasingly quiet submarines, Air-Independent Propulsion[d] (AIP) reducing indiscretion rates, advanced submarine weaponry with terminal effectiveness and stand-off ranges, and difficult physics of seawater—such as thermoclines[e] and ambient noise (Figure1)[5]—that obscure detection. These developments have enhanced submarines’ offensive roles while complicating ASW planning and execution. Therefore, modern ASW systems must detect stealthy submarines, track them without revealing their own position, communicate securely, and execute precise, reliable strikes. With the advancing sensor technology, digitalisation, data analytics, uncrewed systems, autonomy, and machine-learning, this paper discusses the scope of Cognitive Digital Twins (CDTs) as real-time decision-support tools in strategic CCCI networks, enabling accurate detection and deterrence.
Figure1: Evolution of Submarine and ASW Technologies
Source: Author’s own
The subsea domain has emerged as one of the most consequential theaters of modern conflict. Emphasis on subsea warfare stems from regional security concerns related to national sovereignty, territorial integrity, and the need to protect sea lines of communication (SLOC). The increasing capability of modern submarines is evident from the Virginia, Yasen-M, and Shang-class, which can strike targets thousands of miles away while operating stealthily.[6] The Columbia-class is currently considered the stealthiest and most advanced, featuring anechoic coating,[f] pump-jet propulsors,[g] and a Large Aperture Bow (LAB) sonar system.[h]
A core ASW objective is to quickly detect, localise and eliminate hostile submarines. In vast open-ocean spaces, triangulation and lateration for precise localisation are particularly challenging and involve ships, helicopters, and deployable sonobuoys (SBs).[i] ASW aircraft and helicopters extend a ship’s situational awareness well beyond the horizon, serving as force multipliers.[j] Airborne ASW have the advantages of greater range, persistence, and coverage. It typically begins at the datum point—the last known location of a target—and expands as helicopters and aircraft equipped with dipping sonar and SBs continue the search (Figure 2). ASW becomes more complex when Target Submarine (TS) employs evasive strategies.[7] The P-8A Poseidon submarine hunter aircraft, equipped with high-resolution sonar and advanced radar systems, remains a cornerstone of ASW.[8]
Figure 2: Locating a Hostile Submarine Deploying Sonobuoys
Source: Author’s own
ASW technologies are increasingly focused on ISR and shifting toward AI-enabled autonomous and unmanned warfare, involving long-range, long-endurance surface and subsea systems with intervention and swarm capabilities. Uncrewed Underwater Vehicles (UUVs) are uniquely suited for ASW information collection with their ability to operate autonomously at long stand-off distances and in shallow waters, offering a level of clandestine capability not possible with other systems.[9]
Figure 3: Long-range Autonomous Defence Vehicles for ISR
Source: Defense Advanced Research Projects Agency[10]
The US Defense Advanced Research Projects Agency (DARPA) has developed two systems under the “Distributed Agile Submarine Hunting (DASH)” programme: the Transformational Reliable Acoustic Path System (TRAPS), a fixed sonar node designed for large-area coverage; and Submarine Hold at Risk (SHARK), an unmanned underwater vehicle (UUV) with active sonar for tracking hostile submarines post-detection (Figure 3a).[11] The US Navy has awarded Boeing a US$274 million contract to develop, test, and deliver five XLUUV Advanced Undersea Prototypes for its Orca programme. The modular Orca XLUUV (Figure 3b) includes guidance and control, navigation, autonomy, situational awareness, communications, power systems, propulsion and manoeuvring, and mission sensors.[12] Boeing’s Echo Voyager extra-large UUV (Figure 3c), weighing ~50tons in air, is host-ship independent with a range of 6,500nm. It features advanced navigation sensors, active buoyancy control system, obstacle avoidance, terrain following, and modular payload capacity. Its hybrid rechargeable power system follows autonomous surfacing to recharge lithium-ion batteries using diesel-powered generators. The system can enable a seabed-to-space network for sharing across satellites, unmanned aerial and surface vehicles, and manned assets, enhancing situational awareness for ASW.[13] Figure 3d shows the Unmanned Surface Vehicle (USV) Sea Hunter, and Figure 3e shows the remotely-piloted UAVMQ-9B (Sky guardian), designed for over-the-horizon ISR missions of 40+ hours in all weather, capable of safe integration into civil airspace, enabling joint forces and civil authorities to deliver real-time situational awareness anywhere in the world, day or night.[14] Figure 4 illustrates technological trends in Autonomous Underwater Vehicles (AUVs) with ranges of thousands of kilometers without recharging.
Figure 4: Long-Range AUVs
Source: Author’s own
ASW planning demands analysing large amount of data, often time-late and ambiguous. Search modelling that incorporates time, TS motion, and the synergistic effects of the SBs is known as the Cumulative Detection Probability (CDP). Computer-aided search (CAS) uses the Monte Carlo (MC) technique for generating multiple notional “tracks”, each representing a potential TS pathway through the search area. Bayesian search locates the TS by analysing a series of events that could have led to its current location. Each permutation of motion is assigned a probability weight based on its likelihood. These are entered into the CAS, which simulates scenarios using the weights to produce a probability distribution function (PDF), or “prior distribution”, indicating areas of higher or lower TS probability.[15] As the ASW search operation progresses, collected information—positive, ambiguous, or negative—is fed into the CAS computer. Like the initial permutations, this data is assigned confidence weights. Bayes’ theorem[k] updates the prior distribution with this new information, producing the posterior distribution (probability map) indicating the TS’s likely location. Using optimal search theory, CAS can produce search plans for a single aircraft, coordinated multiple-aircraft missions, or sequential sorties to maximise the probability of TS detection.[16]
A digital twin (DT) is a dynamic digital representation of an object or system, described through equations that define its characteristics and properties. It spans the entire lifecycle, continuously updated through real-time data streams from Internet of Things-based sensors. DTs employ simulation, machine-learning, outcome analysis, and reasoning to support decision-making. They are valuable for conceptual configuration, understanding products or system behaviour, enhancing design, and monitoring or managing systems in real-time. Depending on the application level, DTs are classified as Component Twins, Asset Twins, and System Twins.[17]
Component Twins provide detailed information about a component’s performance and behaviour, both in real-time and over extended periods. Asset Twins, comprising multiple Component Twins, deliver real-time information on an asset’s operational status, performance data, and environmental conditions—such as in buildings, machines, and vehicles—helping improve safety and operational efficiency. System Twins monitor and analyse overall system performance, identify inefficiencies, and support optimisation. The typical architecture of an AI-enabled, situation-aware predictive DT used for decision support is represented in Figure 5. It includes numerical models of subsystems, sensors generating standardised data streams on operational status, and a real-time interface with a data repository for enabling machine-learning. Potential failures, dynamic conditions, and abnormal environments can be simulated using the DT for decision-making in real-time.
Figure 5: DT in Situation-Aware Decision-Making
Source: Author’s own
The concept of a physical twin was first applied during NASA’s Apollo 13 mission in 1970, when ground engineers had to quickly account for changes to the spacecraft—322,000 km away—under extreme space conditions with lives at stake.[18] Advancements in digital technologies later led to the development of the DT, first described by David Gelernter in his 1991 book Mirror Worlds.[l] The DT concept was first publicly introduced in 2002 by Michael Grieves for product lifecycle management, and later used by John Vickers of NASA in a 2010 roadmap report. Although the DT concept has been familiar since 2002, its application across various domains began only after 2017.
Presently, in addition to the space and defence domain, DT technology is applied in manufacturing, aviation, robotics, retail, healthcare, and urban planning. During 2020, the DT of the Lockheed Martin Block IIR GPS satellite (SatSim), developed for the US Space and Missile Systems Center (SMC), enabled vulnerability scans and penetration tests across the entire GPS system—including the satellite, ground control station, and the radio frequency link between them.[19]
Extensive use of DT is reported in the space and aviation industries. General Electric, Rolls-Royce, and Pratt and Whitney use AI-based DTs to simulate individual engines at engineering centers using real-time data from in-flight counterparts. This allows accurate assessment of structural component lifespan, improved engine performance, and enhanced operational safety and flexibility.[20] In GE’s landing gear DT uses sensors placed at failure-prone points—such as hydraulic pressure and brake temperature—to provide real-time data for predicting malfunctions and estimating the life-cycle.[21] Incorporating sensor data from real-world vehicles/systems enhances simulation accuracy and helps identify blind spots in virtual test databases.
DTs are used to train AI systems for autonomous aircrafts, enhancing their ability to interpret surroundings and understand algorithms that respond to diverse scenarios. Engineers at Cranfield University are working on a “conscious aircraft”, which involves creating a DT of the entire aircraft by integrating monitoring systems and interpreting the results using AI.[22] Modern aviation maintenance software now integrates DTs to streamline maintenance management.
At Boeing, DTs predict the performance of aircraft components across their lifecycle—starting with low-fidelity models that are refined over time. They are also used to digitise aircraft systems, enable information sharing across the supply chain, and optimise cargo load balance.[23]
In robotics, DTs of robotic cells allow parallel development of mechanical, electrical, and automation design, along with system engineering trials including modelling, kinematics,[m] calibration, and commissioning (such as in ABBIRB-120).[24]
In the automotive domain, DTs use live data to track energy consumption under different driving conditions and weather patterns. While early DT models may be imperfect, comparing outputs with real-world data helps improve vehicle design. Formula1 teams are developing DTs for their racing cars using extensive data from testing and races. Each car is equipped with 150-200 sensors that transmit high-accuracy, low-latency data every millisecond to their tech centres.[25] Recently, Ford introduced DT-enabled predictive headlights to enhance night-time visibility.[26] In the power sector, DTs are being developed for real-time grid analysis, integrating historical and current data to inform producers and consumers of the current grid status and forecast future performance. DTs are also used in renewable energy systems, including solar and wind power.[27]
In the transportation and logistics segment, SITA Lab is developing a fully functional DT of a US East Coast airport. During disruptions, the DT can simulate scenarios involving aircraft arrivals and departures, passenger volumes, queue wait times, escalator operations, restroom satisfaction, and traffic flow at drop-off and pick-up points.[28] Similarly, DTs of smart cities are used to simulate and analyse different scenarios such as traffic flow, energy consumption, optimisation of traffic signal timing, and emergency response.[29]
In product development, DTs are used to create concept configurator during the development phase—from mission requirements and ConOps to customer requirement analysis. In high-tech manufacturing, DTs help identify emerging issues and simulate the effects of upgrades and design changes, reducing the need for extensive field tests. In maintenance and operations, DTs support monitoring, diagnostics, and prognostics[n] to optimise asset performance and utilisation, and offer decision-support during hard-to-predict scenarios. Lifecycle assessment (LCA) twins are used for residual life assessment and major technology upgrades (Figure 6).
Figure 6: DT Used for Production, Operational Support, and LCA
Source: Author’s own
In India’s Samudrayaan-Matsya6000 deep-ocean human scientific submersible developed by Ministry of Earth Sciences- National Institute of Ocean Technology, a first-of-its-kind Cognitive DT (CDT) named Chaitanya acts as a co-pilot, supporting the crew during 15 hard-to-predict life-critical scenarios by generating emergency operating protocols (EOPs) to manage life-support resources for up to 108 hours.[30] CDT Chaitanya (developed along with SRM Institute of Science and Technology, Chennai) with machine-learnt AI-capabilities co-exists with Matsya6000, updated with essential sensory information in real-time, simulates scenarios “ahead-of-time”. It is based on multiple complex multi-disciplinary coupled models involving interactions between human physiology and engineering equipment under different operating modes in varying subsea environmental conditions. Chaitanya also provides information about the health of the subsystems, their effect on the other subsystems, and on Matsya6000 as a whole, with reasoning abilities (using reinforced learning) to predict potential failures before they become critical, thereby enabling self-consciousness (Figure 7 and 8).
Figure 7: Cognitive Digital Twin Chaitanya Used in Matsya6000
Source: Author’s own
Figure 8: CDT Chaitanya as a Co-pilot of Matsya6000 Supporting Life-critical Scenarios
Source: Author’s own
Advances in CDT capabilities are being driven by four key technologies: the Internet-of-Things (IOT) and Big data, advanced analytics (machine-learning), cloud-based computing Power, and accessibility by remote devices. A fully functional CDT has demonstrated remarkable results, and its application in the ASW could be a game-changer.
With the advancing autonomous systems, sensor technology, digitalisation, communication, networking, data analytics, and machine-learning, strategic ASW is shifting towards CCCI, also referred to as the Theater ASW Commander (TASWC).[31] AI-enabled DT, using inputs from multiple uncrewed autonomous systems, can support the development of CCCI network with enhanced battle space awareness and cognitive intelligence, enabling precise detection and deterrence capabilities (Figure 9).
CDTs can also help mitigate human error during extended operations and reduce cognitive strain[o] during unforeseen situation or HTPS. During CCCI-based ASW, aircraft (Figure 6a) deploy SBs to extend detection range; Marine Autonomous Surface Ships (MASS) with torpedo/decoy launchers (B) conduct assaults and recharge AUV batteries (C); AUVs with intervention capabilities (D) are used for mine detection and neutralisation; and AUV swarms (E), capable of intercommunication with each other, the MASS, and the CCCI theatre, further extend the mission’s detection range.
Figure 9: Emerging Concept of CCCI Theatre in ASW
Source: Author’s own
The CCCI theatre operating a CDT, based on inputs from land, sea, air, space and cyber systems, helps to run and validate complex decision-making matrix and provide better and more robust strike solutions (Fig.7). It could provide input to operational, tactical, and strategic levels, during a HTPS for experienced commanders. Real-time decision support on ASW using a CDT refers to use of software and systems to help commanders make effective, informed decisions “on-time” during the ASW mission. CDT-aided ASW provides a comprehensive view of the situation and suggest timely optimal decisions. With the increased complexity of technology, data-driven system, better sensors, AI and fast computing power, the CDT can be a force multiplier for naval war-fighting ability.
As indicated in Figure 10, CDT could support multiple phases of ASW including data integration, fusion, provide a tactical picture for target identification, assessment and classification of suspicious target. It could support search region optimization, ASW asset optimisation for planning an assault, as well as execute a tactical manoeuvre. A CDT could offer a significant advantage in ASW by providing a virtual, real-time representation of ASW platforms, allowing for improved sensor control, simulation, and training. They enable dynamic interaction between the simulation world and the real world, facilitating enhanced decision-making and optimised resource allocation.
Figure 10: CDT for Effective ASW
Source: Author’s own
A typical architecture of a decision-support CDT for ASW is shown in Fig.11. The inputs to CDT include range and heading from the sonobuoys (SBs) (SB 1, 2 and 3) that are initially deployed (from the ASW aircraft) in the suspected search location, environmental information, location bathymetry, sound velocity profile (SVP) and the geo-coordinates of the ASW ship and aircraft. The SBs provide its range and bearing from the target submarine (TS) estimated based on Received Signal Strength (RSS) and Time Delay Estimation (TDE), respectively. Based on these inputs the CDT computes the trajectory of the target submarine (TS).[32]
Figure 11: CDT for CCCI ASW Application
Source: Author’s own
The objective of the CDT is to continuously track the TS and carryout an early encounter (Figure 11). As the TS moves away from the SB, the RSS decreases and hence ineffective for the localisation. Hence additional SB (SB4, SB5…SBn) needs to be deployed at appropriate locations. For identifying appropriate locations of deployment of subsequent SBs, the CDT needs to carry out ocean acoustic signal propagation modelling and simulation (using Bellhop, DESERT, SUNSET and NS2 software) with real-time inputs from operating ocean environment, sound velocity profile, seafloor bathymetry, TS speed, and its anticipated trajectory based on time-since-detection.[33] At the same time, the CDT could determine the optimum location and attitude of the ASW ship and the kinematics of ASW ship’s torpedo launcher platform for executing a precise encounter.
The DT is a comprehensive mathematical model or virtual replica of the MASS/AUV, capturing system-specific dynamics, power, propulsion, navigation, positioning, ballast, dynamic positioning, and other automation systems. It co-exists with its physical counterpart, mapping dynamic behaviour in real-time and simulating ocean conditions using inputs from environmental, navigation, and obstacle-detection sensors—enabling efficient control of the platform.
The DT integrates sensor data from the onboard Integrated Vehicle Health Management (IVHM) System, maintenance records, and historical and fleet-wide data obtained through data mining. By combining these inputs, the CDT continuously forecasts the vessel’s or subsystem’s health, remaining useful life (RUL), and the probability of failures over time.
Figure 12: DT for Manned and Unmanned Ships
Source: Digital Twin, DNV[34] Vedachalam et al.[35]
The DT implementation for manned ships and Marine Autonomous Surface Ships (MASS) is illustrated in Figure 12. In manned ships, DTs are operated onboard, whereas for MASS/AUV platforms, the CDT can be hosted at the shore-based CCCI theater, with sensor data streamed in real-time via offshore satellite communication networks. Figure 13 shows the configuration of a shore-based MASS CDT, where data—including ship kinetics and kinematics—from the offshore MASS are transmitted to the CCCI-CDT through satellite communication networks.
Figure 13: Configuration of MASS Involving Shore-Based CDT
Source: Author’s own
The CDT based on the numerical or ML data-driven models and software-based control algorithms, uses subsystems streamed data to enable time-domain analysis system and visualisation. It can simulate vessel or subsystem behaviour under various degradation or failure scenarios (Figure 14). The CDT’s outputs are critical for assessing and maintaining vessel reliability and performance during ASW operations—particularly by emulating safety-critical and HTPS under both normal and degraded conditions.
Figure 14: CDT Support in MASS and AUV Functions
Source: International submarine engineering limited [36]
Advanced sensor technology, vessel digitalisation, data analysis, and autonomous mission or path planning algorithms based on situational awareness are essential for deploying MASS and AUVs in ASW missions. Figure 12 highlights key areas where CDTs are particularly valuable—predicting MASS/AUV responses to safety-critical events in real-time and identifying potential issues by comparing predicted and actual behaviour before they become critical.
Figure 15 shows the coordinate systems used to analyse ship or AUV manoeuvrability. The heading angle (ψ) is the angle between the direction of the x0-axis and the x-axis. In the earth-fixed coordinate system, the ship’s center of gravity represents its position, and the heading angle (ψ) is determined by the orientation of the MASS or UUV.[37]
Figure 15: Coordinates of MASS/AUV Manoeuvrability
Source: Author’s own
Predicting the manoeuvrability of a MASS/UUV using a CDT could be done using a numerical model with real-time inputs on rudder angle, speed/thrust of the main and azimuth propellers,[p] and the sea state. These inputs are essential for implementing navigation protocols. Through supervised and unsupervised machine learning (ML), the CDT learns from numerical simulations and real-world data collected under various operational and environmental scenarios—both normal and degraded.
This learned CDT supports reliable performance during safety-critical situations arising from degraded operations, unpredictable scenarios, and rough ocean conditions in ASW operations.
A key challenge in AUV operations for ASW and other missions is the limited mission duration, constrained by onboard energy and data storage capacity. To overcome this, underwater Homing and Docking (H&D) stations connected to the MASS are being developed. These stations enable battery recharging, mission data upload, and mission profile download, thereby extending subsea ASW mission duration.[38] The MASS-based homing guidance system must reliably guide the AUV back to its dynamic dock, accounting for the AUV’s dynamic response capabilities and residual battery energy.[39]
For an approaching AUV (Figure 16), long-range homing guidance—usually over a few kilometres to several tens of metres—is achieved using proven acoustic baseline systems. These systems involve an acoustic transceiver at the MASS docking station and a transponder on the AUV. The AUV computes its range and bearing relative to the dock to perform course and attitude corrections.
As the AUV nears the dock, precise attitude and pose corrections become critical for successful docking. This requires high-frequency spatial measurements with adequate temporal resolution. An on-board CDT, operating a kinematics model with real-time inputs—such as dock attitude, range, speed, and environmental conditions—can provide corrective attitude inputs and redefine the AUV’s trajectory. This enhances the reliability of homing and docking, particularly under dynamic docking conditions (Figure 14).
Figure 16: CDT for AUV Homing Implementation
Source: ISE[40] and Author’s own
Intervention autonomy refers to the AUV’s ability to perform successful subsea interventions using onboard manipulators. In ASW missions, the Intervention-AUV (I-AUVs) are deployed for tasks such as mine detection, neutralisation, and other countermeasures. Recent developments in this field include I-AUVs Girona500 and Cuttlefish (Figure 17).[41]
Figure 17: Recently Reported I-AUV Developments
Source: Vedachalam et al.[42]
These promising I-UAV technologies requiring high levels of decision-making and navigation autonomy, are currently at level 3 of a 9-level development cycle (with level 3 is successful demonstration in a controlled ocean environment and level 9 indicates routine operational use). These technologies are being prepared for development across multiple domains, including ASW. The I-AUV, programmed with a mission plan and target coordinates, approaches the subsea (e.g., a mine) using a conventional and proven navigation system.
Successful subsea intervention requires close coordination between the AUV and its robotic manipulator, both of which have limited degrees of freedom (DoF). Tight, real-time coupling between the two enhances the I-AUV’s dexterity in mine operations. When in proximity to the target, an onboard CDT can enable precise intervention for mine detection and neutralisation.
Figure 18: CDT for Autonomous Subsea Intervention
Source: Author’s own
A typical CDT (Figure 18) can use onboard cameras and image processing algorithms to determine its pose relative to a task and calculate the distance between the task and the manipulator’s end-effector. These parameters, along with vehicle attitude data from the navigation sensor suite, are fed into a trained machine learning model. The model then computes the Denavit–Hartenberg (D-H) parameters[q]required for the manipulator joints and vehicle actuators to achieve precise positioning and actuation of the end effector.[43]
For a particular scenario, the CDT can account for the limitations in the DoF of the manipulator and vehicle, thus ensuring precise intervention with the subsea target.
Subsea Swarm Robotics Systems (SRS) is a system of multiple AUVs (inspired from the spatial self-organising, navigation behaviour and collective decision-making behaviour of social animals) that helps in realising efficient, robust, scalable, and flexible multi-AUV configurations (Figure19).[44] SRS is used in ASW missions to conduct rapid surveys of mines or suspicious targets.
A typical CDT architecture for multi-AUV SRS implementation is shown in Figure 20. The CDT, operating in the deployment vessel, receives inputs such as the survey map, real-time positions of the deployment ship and swarm AUVs, environmental information, bathymetry, Sound Velocity Profile (SVP), sonar parameters (including acoustic power[r] level, frequency, and swath[s]), and the desired altitude (height from the seafloor) of the AUV.
Figure 19: UUVs Operated in Swarm Mode
Source: Author’s own
Based on input data, the CDT generates an integrated swarm mission survey line plan and provides trajectories for individual AUVs. These are processed by each AUV’s mission controller for path planning). The CDT also determines the optimal positioning and trajectory of the deployment platform relative to the swarm.
An underwater acoustic signal propagation model assists the CDT in near real-time to establish operational boundaries for the AUVs, based on sonar parameters, SVP, and environmental conditions. Moreover, the CDT tracks completed and remaining survey areas, estimates mission completion time, and monitors residual battery life for each AUV to facilitate timely retrieval. By doing so, the CDT supports advanced decision-making for resource planning—such as adjusting the number of AUVs, determining the required speed and altitude, and meeting preferred mission durations.
Figure 20: CDT for Swarm AUV Implementation
Source: Author’s own
Advanced subsea warfare technologies are under development, including modern stealth submarines equipped with Air Independent Propulsion and advanced combat systems featuring torpedoes, mines, and submerged-launch missiles. As these submarines gain greater stealth and endurance, ASW will remain a core naval mission. However, countering future undersea threats will become increasingly complex, as conventional ASW approaches may no longer provide sufficient situational awareness or tactical control.
To overcome these challenges, advanced technologies and new operational approaches—such as distributed, deployable/off-board ASW sensor networks with autonomous mine countermeasure capabilities—are being developed. ASW is expected to shift from “platform-intensive” to “sensor-rich” operations, driven by the collection and fusion of data from all-source intelligence, surface and underwater assets, aerial assets, autonomous vehicles, space-based asset systems, and seabed sensor arrays.
Amidst the growing trend of networking unmanned ASW assets, a situation-aware CDT could function as a real-time decision support system within CCCI. Such a CDT would significantly enhance ASW capabilities by expanding battle-space awareness and improving the detection, tracking, classification, and neutralisation of underwater threats with greater accuracy and reliability—especially by mitigating risks associated with human cognitive limitations.
[a] It is the ability of a nuclear-armed submarine to survive a first strike from an adversary and then launch a retaliatory nuclear attack.
[b]System for long-range detection, tracking, and localisation of submarines, comprising a long string of passive hydrophones towed behind a ship or submarine to detect underwater sounds.
[c] Employing unconventional tactics to exploit the vulnerabilities on the opponent’s side.
[d] It is a technology that allows non-nuclear submarines to operate underwater for extended periods without needing to surface or use a snorkel to access atmospheric oxygen.
[e]layer in the ocean where the temperature changes rapidly with depth, which acts as a barrier to sonar because sound waves bend or refract as they pass through the temperature gradient. Submarines can use the thermocline to their advantage by positioning themselves within or below it, making them harder to detect with active sonar.
[f]This coating helps absorb sonar waves, reducing the submarine's acoustic signature and making it harder to detect.
[g]This design reduces noise compared to traditional propellers, further enhancing stealth.
[h]A sophisticated sonar system offers enhanced detection capabilities while minimising its own acoustic footprint.
[i]Expendable, air-droppable devices to detect, classify, and track underwater sounds, primarily from submarines. It transmits acoustic data back to an aircraft or surface vessel for analysis.
[j] Capabilities or attributes that significantly enhance the combat effectiveness of a naval force with the same resources.
[k] Bayes' theorem is used to update the probability for target submarine detection based on more evidence, as and when the information becomes available.
[l] It lays out the foundational idea of creating digital replicas of real-world entities.
[m] Deals with the geometric aspects of motion, such as position, velocity, and acceleration, and how these quantities relate to each other.
[n] Engineering method of predicting its future performance and failure behaviour.
[o] It is temporary state where the commanders’ cognitive abilities, such as thinking, memory, and decision-making, are impaired, leading to human error.
[p] A part of marine propulsion system that provides both thrust and steering capabilities by rotating a propeller mounted on a pod 360 degrees horizontally.
[q] They are a systematic method to represent the kinematic chains of robotic arms by providing a standard notation to describe the relative positions and orientations of adjacent links.
[r] It is the measure of the total sound energy produced by a source per unit of time.
[s]It's the width of the area being mapped or measured at a given time.
[1]A.K. Chawla, “Unseen & Unheard: The Role of Stealth,” SP's Naval Forces May 2023, https://www.spsnavalforces.com/story/?id=842&h=Unseen-and-Unheard-The-Role-of-Stealth
[2]Robert F Hendrik, “Surtass Twinline,” Johns Hopkins APL Technical Digest 34 (2018): 179-193.
[3]Yashodhan Mirwankar, “Comparative Analysis of the Maritime ISR Capabilities of India and China in the IOR,” Electronic Journal of Social and Strategic Studie 5 (2024): 295-329.
[4]Ronald O’ Rourke, Navy Large Unmanned Surface and Undersea Vehicles: Background and Issues For Congress, Congressional Research Service, 2025, https://sgp.fas.org/crs/weapons/R45757.pdf
[5]Tytti Erästö, Fei Su and Wilfred Wan, Navigating Security Dilemmas in Indo-Pacific Waters: Undersea Capabilities and Armament Dynamics (Stockholm: SIPRI Publications, 2024), pp. 18.
[6]Russia Submarine Capabilities Fact Sheet August 28, 2024.
[7]E.R. Van Veldhoven and H.J. Fitski, “A Priori Planning of ASW Operations: Providing a Robust Mission Advice,” Proceedings of the NATO STO SET-244 Symposium, 2017.
[8]U.S. Director, Operational Test & Evaluation (DOT&E) Annual report 2013, Navy Programs : P‑8A Poseidon Multi‑Mission Maritime Aircraft, https://www.dote.osd.mil/Annual-Reports/2013-Annual-Report/
[9] “Distributed Agile Submarine Hunting (DASH),” https://www.darpa.mil/research/programs/distributed-agile-submarine-hunting
[10] “Defense Advanced Research Projects Agency (DARPA),” https://www.darpa.mil/
[11] “Distributed Agile Submarine Hunting (DASH),” https://www.darpa.mil/research/programs/distributed-agile-submarine-hunting
[12]Nico Palmroos, “Extra-Large Unmanned Underwater Vehicles (XLUUVs): Payload Benefits, Technological Advancements and Military Utility in the Baltic Sea” (Master thesis, Swedish Defence University, 2023), pp 2-31.
[13] “echo_voyager_product_sheet,” https://www.boeing.com/content/dam/boeing/boeingdotcom/defense/autonomous-systems/echo-voyager/echo_voyager_product_sheet.pdf
[14] “US Navy Showcases Sea Hunter Unmanned Surface Vehicle at LA Fleet Week,” NAVSEA News, June 14, 2024, https://www.navsea.navy.mil/Media/News/Article-View/Article/3808190/us-navy-showcases-sea-hunter-unmanned-surface-vehicle-at-la-fleet-week/
[15]Yongzhao Yan et al., “Research on Anti-Submarine Warfare Method of Unmanned Aerial Vehicle Cluster Based on Area Coverage and Distributed Optimization Control,” Drones 8, no. 12 (2024): 2504-446X.
[16]Yongqin Sun et al., “A Intelligence Target Situation Assessment Method for Cooperative Search by ASW Patrol Aircraft” (paper presented at 14th International Conference on Machine Learning and Computing, Shenzhen, China, February 18-21, 2022).
[17]Angira Sharma et al., “Digital Twins: State of the Art Theory and Practice, Challenges, and Open Research Questions,” Journal of Industrial Information Integration 30 (2022): 100383.
[18] B.D. Allen, “Digital Twins and Living Models at NASA” (paper presented at Digital Twin Summit, Milford, USA, November 3-4, 2021).
[19]AFCEA International Signal Media, “Digital Twinning Takes Flight ,” https://www.afcea.org/signal-media/digital-twinning-takes-flight
[20]Rolls-Royce, “Digital Twin,” https://www.rolls-royce.com/innovation/digital/digital-twin.aspx.
[21]Patrick K. Malone, “Economics of Digital Twins in Aerospace and Defense” (paper presented at the ICEAA Professional Development & Training Workshop, Minneapolis, USA, May 14-16, 2024).
[22]Cranfield University, “Conscious aircraft,” https://www.cranfield.ac.uk/centres/digital-aviation-research-and-technology-centre/conscious-aircraft
[23]NEC, “The Future of Digital Twins,” https://www.nec.com/en/global/insights/article/2020022510/index.html.
[24]R.A.P. Vitalli and J. M. Rosário, “Digital Twin of ABB IRB-120 Robotic Cell Based on the Industry 4.0,” International Robotics & Automation Journal 8, no. 3 (2022): 89-92.
[25]Loffredo Gabriel, “Digital Twins: Revolutionizing The Development Of Formula 1 Cars With AI And Cloud,” Globant Medium, July 18, 2024, https://medium.com/globant/digital-twins-revolutionizing-the-development-of-formula-1-cars-with-ai-and-cloud-815a65466c05
[26] William Payne, “Ford Digital Twin for Predictive Headlights,” IMC IoT M2M Council, October 27, 2021, https://www.iotm2mcouncil.org/iot-library/news/connected-transportation-news/ford-digital-twin-for-predictive-headlights/
[27]J.V.S. Amaral et al., “Energy Digital Twin Applications: A Review,” Renewable and Sustainable Energy Reviews 188 (2023): 113891.
[28]Kevin O'Sullivan, “Digital Twins, the Airport Operations Control Interface of the Future,” SITA, April 11, 2019, https://www.sita.aero/pressroom/blog/digital-twins-the-airport-operations-control-interface-of-the-future/
[29]Bronislava Halúskova, “Digital Twin in Smart City,” Transportation Research Procedia 74 (2023): 1471–1478.
[30]Sahana Gosh, “India Tests First Crewed Deep-sea Submersible-Successful Wet Trials Ready Matsya 6000 for 2026 Scientific Mission,” Nature, February 25, 2025, https://www.nature.com/articles/d44151-025-00032-2
[31]“Creation of Theatre Commands will Kick-start Next Generation of Reforms: CDS Gen,” The Hindu, May 24, 2025, https://www.thehindu.com/news/national/indian-armed-forces-restructuring-theatre-commanders-to-look-after-force-application-service-chiefs-force-generation/article69609933.ece
[32]J. Yang, “A Digital Twins Enabled Underwater Intelligent Internet Vehicle Path Planning System via Reinforcement Learning and Edge Computing,” Digital Communications and Networks10, no. 2 (2024): 282-291.
[33] S. Wynn Simmonds, “Comparative Analysis of Underwater Acoustic Propagation Models: Evaluating the Trade-off Between Accuracy and Computational Demands for Coastal Noise Predictions,” Marine Pollution Bulletin 215, 2025: 117905.
[34] Margrethe Andersen, “HD Hyundai Receives DNV AiP for Cloud-based HiDTS Digital Twin System,” DNV.com, https://www.dnv.com/news/2024/hd-hyundai-receives-dnv-aip-for-cloud-based-hidts-digital-twin-system/
[35] Narayanaswamy Vedachalam, “Assessment of the Technological Maturity of Marine Autonomous Surface Ships,” Marine Technology Society Journal 56, no. 4 (2022): 43-58.
[36] International Submarine Engineering Limited, “Autonomous Docking,” https://ise.bc.ca/product/dock/
[37] H.S. Berg et al., “Digital Twin Syncing for Autonomous Surface Vessels Using Reinforcement Learning and Nonlinear Model Predictive Control,” Scientific Reports 15, no.1 (2025): 9344.
[38] BN Vandavasi et al., “AI‐ML-Enabled Electromagnetic Homing Guidance System for Scientific Autonomous Underwater Vehicles,” Marine Technology Society Journal 57, no. 1 (2023): 7-15.
[39]International Submarine Engineering, “Autonomous Docking,” https://ise.bc.ca/product/dock/
[40] Narayanaswamy Vedachalam, “Design Considerations for Strategic Autonomous Underwater Swarm Robotic Systems,” Marine Technology Society Journal 54, no. 2 (2020): 25-34.
[41]T. Slawik et al., “Attitude Control of the Hydrobatic Intervention AUV Cuttlefish using Incremental Nonlinear Dynamic Inversion” (paper presented at IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Abu Dhabi, United Arab Emirates, October 14-18, 2024.
[42] Narayanaswamy Vedachalam et al., “Design Considerations for strategic Autonomous Underwater Swarm Robotic Systems,” Marine Technology Society Journal 54, no. 2 (2020): 25-34.
[43] D.Calandra et al., “Digital Twin-and Extended Reality-based Telepresence for Collaborative Robot Programming in the 6G Perspective,” Digital Communications and Networks 10, no. 2 (2024): 315-327.
[44] Pollyanna G. Dias et al., “Swarm Robotics: A Perspective on the Latest Reviewed Concepts and Applications,” Sensors 21, no. 6 (2021): 2062.
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.
Dr. N. Vedachalam is Senior Scientist and Program Director at the National Institute of Ocean Technology an autonomous ocean research centre under the Ministry of ...
Read More +
PREV
