Monitoring marine environments is essential for understanding ecological dynamics and preserving underwater sites of scientific and cultural interest. While autonomous robotic systems can cover large areas and collect high-quality water data, scientific divers remain indispensable for survey operations due to their direct observations and selective sampling. However, underwater navigation presents significant challenges for researchers, including poor visibility, marine currents, and the absence of GNSS signals. To address these challenges, this study proposes the conceptual design of a motorized smart buoy capable of autonomous environmental monitoring and providing operational support to divers. The system integrates a suite of sensors for real-time measurement of physicochemical water parameters, a GNSS module for surface localization, and an acoustic positioning system to enhance underwater navigation. This system determines the diver’s position by emitting and receiving acoustic signals. Once the position is calculated, the information is transmitted directly to the diver’s tablet, enabling real-time awareness of their location, improving navigational safety, and allowing the collection of georeferenced data for more accurate environmental analysis. Furthermore, the implementation of Deep Reinforcement Learning (DRL) algorithms has been considered to optimize the trajectory of the smart buoy in dynamic marine environments, enhancing energy efficiency and monitoring accuracy. The proposed system represents a significant advancement in marine ecosystem monitoring, combining operational autonomy with adaptability to diverse application scenarios.