PEP LLUÍS NEGRE

Leading the technical direction of Rosenxt's autonomous subsea robotics programme for close-range asset inspection, with a current focus on physical AI — agentic decision-making, learned behaviours, and semantic perception — to turn a hard underwater autonomy problem into a product that inspects critical infrastructure reliably, at depth, without a human at the controls. Setting technical strategy: choosing the AI, autonomy, and perception approaches that take the programme from prototype to fielded system. Working closely with product and stakeholders so that requirements and engineering reality meet at the right point. Leading a cross-functional team across AI, robotics, simulation, software, DevOps and R&D operations. Also responsible for how the team scales and hires, the processes that keep a growing team fast, and the budget requests and project documentation that keep the programme funded and legible to the business.

Delivered production-grade robotics software including SubSLAM's real-time 3D stereo-vision system and the world-first subsea autonomous inspection algorithm, combining vision, mapping and ML components. Built custom pipelines for point cloud processing, dense 3D reconstruction, and SLAM robust to visually degraded or GPS-denied environments; co-inventor on related patents.

Led the development of world-class AI-driven subsea autonomy. Technically led and delivered the world-first autonomous inspection of an offshore jacket structure. Drove the end-to-end architecture and algorithmic design — from perception and sensor fusion to decision-making and mission autonomy. Led a multidisciplinary tech team, set technical direction, and ensured delivery of high-impact, production-grade solutions.

Researched and prototyped emerging computer vision technologies to guide subsea exploration and inspection. Designed and developed the core of Rovco's cutting-edge live 3D vision for ROVs and AUVs. Tackled challenging underwater imagery problems to keep improving camera pose accuracy and dense 3D models.

Designed online 3D underwater mapping algorithms, based on an accurate Visual-SLAM positioning system and semi-dense 3D reconstructions. Developed cutting-edge solutions for reliable and impressive 3D underwater reconstructions, executed online during the robot mission. Main contributor and inventor of Rovco's SubSLAM online 3D mapping sensor for subsea infrastructure. Delivered the most advanced subsea stereo camera system to date.

Co-responsible for the entire platform (software and hardware) of the autonomous underwater robot "Turbot". In charge of the whole localization pipeline based on various sensors: IMU, DVL, USBL, GPS and two stereo cameras. Specialized in stereo localization, loop-closing detection, and 3D reconstruction of the environment, along with the tools to visualize it. Responsible for the design and development of the vehicle's operating interface, based on HTML5, the Google Maps API, and Rosbridge.

Led a platform to write and distribute books online. In charge of the website and apps developed for iOS and Android. Created value, managed the team, and kept developing code in HTML5, JavaScript, CSS, PHP, Node.js, Xcode and Java.

Developed drivers for IAS/Wonderware, cameras, relays, and many other sensors for factories and intelligent houses. Implemented power monitoring systems to improve building efficiency with robust web front-ends and databases. Responsible for the project to automate the reading of water and electricity meters for the Port of Palma de Mallorca.

Responsible for the control design and tuning of the ECO74 and ECO80 wind turbines. Specialized in system identification and fault-tolerant control in adverse weather conditions.

Advanced control techniques. Mobile robotics. Computer vision.

Computer vision. Robotics. Electronics. Control theory.