NemeSys: An Online Underwater Explorer with Goal-Driven Adaptive Autonomy
Adnan Abdullah, Alankrit Gupta, Vaishnav Ramesh, Shivali Patel, Md Jahidul Islam RoboPI laboratory, Department of ECE, University of Florida, FL 32611, USA,Email: {adnanabdullah@, gupta.alankrit@, vaishnavramesh@, spatel30@, jahid@ece.}ufl.edu · 2025
Read NemeSys: An Online Underwater Explorer with Goal-Driven Adaptive Autonomy by Adnan Abdullah, Alankrit Gupta, Vaishnav Ramesh, Shivali Patel, Md Jahidul Islam RoboPI laboratory, Department of ECE, University of Florida, FL 32611, USA,Email: {adnanabdullah@, gupta.alankrit@, vaishnavramesh@, spatel30@, jahid@ece.}ufl.edu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-11889.txt.
Introduction
Adnan Abdullah and co-authors build an underwater robot that takes new orders mid mission through a faint magnetic electric link, in work from twenty twenty-five. Ocean robots work without GPS and with weak contact, so most run fixed scripts. Changing course mid dive is rare and slow.
Their system pairs a mission code format with a control stack built for quick retasking. A digital twin tested task switches in under fifty milliseconds with lean computing overhead. Tank tests and open water trials confirmed steady control and clean mission runs. The team frames the robot as a step toward responsive, goal driven exploration with a remote operator in the loop.
Catalog record
| CREATOR | Adnan Abdullah, Alankrit Gupta, Vaishnav Ramesh, Shivali Patel, Md Jahidul Islam RoboPI laboratory, Department of ECE, University of Florida, FL 32611, USA,Email: {adnanabdullah@, gupta.alankrit@, vaishnavramesh@, spatel30@, jahid@ece.}ufl.edu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-11889
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
NemeSys: An Online Underwater Explorer with Goal-Driven Adaptive Autonomy Thanks: Thanks: This research is supported by the U.S. National Science Foundation (NSF) grants # 2330416 2330416 and the University of Florida Research grant # 132763 132763 . Adnan Abdullah, Alankrit Gupta, Vaishnav Ramesh, Shivali Patel, Md Jahidul Islam RoboPI laboratory, Department of ECE, University of Florida, FL 32611, USA,Email: {adnanabdullah@, gupta.alankrit@, vaishnavramesh@, spatel30@, jahid@ece.}ufl.edu Abstract Adaptive mission control and dynamic parameter reconfiguration are essential for autonomous underwater vehicles (AUVs) operating in GPS-denied, communication-limited marine environments. However, most current AUV platforms execute static, pre-programmed missions or rely on tethered connections and high-latency acoustic channels for mid-mission updates, significantly limiting their adaptability and responsiveness. In this paper, we introduce NemeSys , a novel AUV system designed to support real-time mission reconfiguration through compact optical and magnetoelectric (OME) signaling facilitated by floating buoys. We present the full system design, control architecture, and a semantic mission encoding framework that enables interactive exploration and task adaptation via low-bandwidth communication. The proposed system is validated through analytical modeling, controlled experimental evaluations, and open-water trials. Results confirm the feasibility of online mission adaptation and semantic task updates, highlighting NemeSys as an online AUV platform for goal-driven adaptive autonomy in dynamic and uncertain underwater environments. I Introduction Autonomous Underwater Vehicles (AUVs) play a pivotal role in oceanographic research, subsea infrastructure inspection, and naval operations [ 1 , 2 ] . Yet, most current AUV platforms operate under a limited autonomy paradigm, executing pre-programmed missions with minimal adaptation to environmental variability, unexpected events, or evolving mission objectives [ 3 , 4 ] . Adaptive mission control, where a human operator can interactively guide or adjust the robot’s behavior during execution, offers a compelling solution to these limitations [ 5 , 6 ] . Fig. 1 : A few snapshots of the NemeSys AUV system operating in the wild; major components are annotated on the right. NemeSys contributes to a new class of AUVs that can offer online mission parameter updates from remote OME signals. Recent advances in terrestrial and aerial domains have demonstrated the power of integrating human inputs with autonomous system execution to improve safety and mission outcomes [ 7 , 8 ] . However, the underwater domain inherently limits real-time communication as electromagnetic waves attenuate rapidly in seawater [ 9 ] . Acoustic communication offers long-range capability but is constrained by limited bandwidth (typically < 10 Kbps), long propagation delays, and multipath effects [ 10 ] . Optical communication can provide higher data rates but is highly sensitive to water turbidity and limited to short-range, line-of-sight conditions [ 11 ] . Consequently, existing AUV platforms are either constrained to fully autonomous missions with little to no adaptivity or require continuous operator oversight through tethered connections or acoustic modems, limiting their scalability and flexibility [ 12 , 13 , 14 ] . Contemporary researchers have demonstrated promising advances in multi-agent relative localization and coordination involving robots, buoys, and navigation markers [ 15 , 16 , 17 , 18 ] . Our prior has developed BlueME [ 11 ] , a novel low-power antenna system that leverages the natural mechanical resonance of compact MagnetoElectric (ME) materials. BlueME enables real-time subsea robot-to-robot coordination at data rates of up to 36 Kbps over distances up to 730 meters, and remains unaffected by line-of-sight or multipath issues. With further integration of green lasers, the Optical-Magnetoelectric (OME) buoys help resolve the directional ambiguity of ME measurements for robust communication and coordination. This makes it an ideal candidate for encoding compact mission updates and transmitting them during untethered AUV operation. In this paper, we present the design and development of a new underwater robotics platform: NemeSys , designed to support untethered mission coordination and adaptive task execution by low-bandwidth OME signaling. To enable goal-driven autonomy and dynamic re-tasking, we develop a lightweight mission encoding framework that translates high-level operator intentions into compact digital commands suitable for low-bandwidth underwater transmission. The operator’s instruction is abstracted into a waypoint pattern such as spiral, grid search, or perimeter scan – and then associated with a bounded set of parameters such as speed, depth, and radius. These dynamic parameters are then encoded into a binary packet using BCH (Bose-Chaudhuri-Hocquenghem) [ 19 ] error correction scheme for updating the subsequent goals of the AUV. These digital packets are transmitted via untethered communication from remote OME buoys. Please refer to our BlueME [ 11 ] paper for details on the communication aspects. In this paper, we focus on the NemeSys system design and its adaptive mission encoding capabilities. Specifically, we demonstrate how the proposed NemeSys AUV reconfigures its trajectory, updates task objectives, and responds to operator interventions from the received mid-mission updates. We conduct extensive evaluations of the NemeSys system, beginning with a comparative analysis of three choices of design configurations. Each design is assessed based on the AUV’s response to external disturbances to characterize its inherent dynamic stability. The configurations are evaluated in terms of maneuverability by empirically determining their heave and yaw rates. We also analyze the AUV’s active control performance through depth regulation experiments using a proportional-derivative (PD) controller. Moreover, we evaluate the mission encoding framework through analytical schemes to balance the error correction capability with encoding efficiency for low-bandwidth underwater transmission. Finally, we validate the platform in open-water environments, where it executes a range of autonomous mission patterns from encoded mission commands. Despite real-world challenges such as strong currents and turbidity, the system demonstrates robust control and successful mid-mission reconfiguration– validating its effectiveness in unstructured underwater settings. Overall, the main contributions are summarized as follows: 1. System design and analyses: We present the design and development of NemeSys AUV , highlighting comparative analysis of multiple configurations in hydrodynamic stability and actuation-induced maneuverability. 2. Mission encoding and control architecture: We introduce a low-bandwidth encoding framework for online mission reconfiguration from remote ME/OME signals. The encoded missions interface with a ROS-based middleware for low-level execution and dynamic trajectory updates. 3. Experimental validations: We perform comprehensive evaluations in laboratory testbeds and by open-water trials to demonstrate robust control and mission execution capabilities of the AUV. 4. Unique features: While some existing AUVs, such as SUNFISH [ 20 ] , CUREE [ 21 ] , and UX1 [ 22 ] offer a broad range of capabilities, NemeSys offers three unique advantages: ( a ) a small footprint: one person can carry and deploy it; ( b ) OME-based dynamic mission reconfiguration and adaptive task execution capability; and ( c ) a low-cost, low-power design with 5+ hours of endurance. Fig. 2 : System design and computational components of the NemeSys AUV: (a) The physical robot; (b) Isometric view of the corresponding CAD design showing overall structural layout; (c) Exploded view highlighting the key components; (d) Top and side views, showing the outer dimensions; and (e) Our ME antenna system from prior work [ 11 ] . II Related Work: AUV Systems for Autonomous Underwater Exploration Classical architectures for AUV autonomy have emphasized robust navigation, environmental mapping, and mission execution in GPS-denied, communication-limited environments [ 23 , 24 ] . Torpedo-like architectures such as the REMUS series [ 25 , 26 ] , Bluefin vehicles [ 27 , 28 ] ,