Curriculum Vitae

November 2024 - Present

As Head of Software, I lead the development of innovative robotics and AI-driven solutions, managing cross-functional teams to deliver high-performance systems.My responsibilities encompass both technical and managerial tasks, such as:

• Designing the software architecture to ensure scalable and reliable solutions.

• Implement CI/CD pipelines.

• Planning and executing innovation projects while driving the ongoing improvement of existing products.

• Managing and mentoring my team to foster collaboration, technical growth, and successful project delivery.

• Contributing to key technical areas, including scanning, path planning, and semantic segmentation.

January 2022 - October 2024

PREEN is revolutionizing the car wash industry with a sustainable, AI-enhanced and robotised technology. Current solution is formed by a scanning system and a 2 robotic arms.

• Semantic Segmentation: Developed car semantic segmentation pipelines using PyTorch on GPU (CUDA), leveraging transfer learning from advanced models like ResNet and SAM to improve accuracy and efficiency.

• Path Planning and Collision Avoidance: Designed and implemented a deterministic 7-axis path planner, coupled with collision detection and avoidance algorithms for safe and efficient navigation.

• 3D Model Reconstruction: Created a 3D car mesh generation algorithm, reconstructing high-fidelity car models from color and depth images.

• Computer Vision and Scene Understanding: Processed camera and depth data for scene interpretation.

• Integration and Deployment: Streamlined code integration and deployment workflows using Docker, Redis, and AWS, ensuring reliability and scalability of software systems.

• CI/CD Practices: Pioneered the adoption of CI/CD workflows with GitHub and AWS, reducing development cycles and enhancing code quality through automated testing and deployment.

June 2021 - December 2021

Company developing equipment for airports. I provide consulting through my company Salazar Technologies. My responsibilities included:

• Implementing an object recognition algorithm by using features and descriptors using PCL and C++.

• Implementing both an offline calibration algorithm for two LiDARs.

September 2019 - December 2020

Start-up company focused on the LiDAR field. Along with the team we built a library in ROS in C++ that could assist the driver of a car in the airport thanks to LiDARs and cameras. My responsabilities included: 

• Software team lead. Planning, designing and implementing the architecture of the code.

• Implementing algorithms in PCL such as RANSAC, ICP and other algorithms like SLAM or LOAM.

September 2018 - August 2019

Consulting company for automotive companies. My responsibilities included:

• Developing Python scripts using the ANSA API for model checking before crash simulations at Porsche.

• Implementing Python graphical interfaces with Qt to manage the database for model checking at Daimler.

September 2019 - August 2021

Activities and Societies: DTU Volleyball Team.

Focus areas: Perception for Autonomous Systems, Software Frameworks for Autonomous Systems, Advanced Autonomous Robots, Deep Learning, Machine Learning, IoT and Intelligent Systems.

February 2018 - August 2018

Exchange semester in Information Technology and Electrical Engineering.

Bachelor thesis in Thoratec.

Focus areas: Introduction to Robotics and Mechatronics, Wearable Systems and Computer Simulation of Sensory Systems.

Septembre 2014 - August 2018

Activities and Societies: UPV Volleyball Team, EUROAVIA and BEST (international engineering student societies).

Focus areas: Mathematics, Physics, Fluid Dynamics, Heat transmission, Circuit Theory, Machine Theory, Economy, Operations Research, Automation Technology, Electronic Technology, Structures, Thermal Machines and Electric Machines.

February 2021 - July 2021

Implementing a novel offline calibration for on-vehicle 2D camera-LiDAR sensor setups by using a spherical target. For this purpose, these tools were used: C++, ROS, OpenCV and PCL. The calibration method was tested in a simulation environment (Gazebo) as well as in real life. The algorithm has an error of around 1.5cm in translation and 0.25º in rotation. 

Check it out in the section Projects! 

September 2020 - January 2021

In this special course, the performance of several 3D sensors were compared to each other. Some of the tests included getting the performance depending on the depth range, material reflectivity, angle of incidence and lightning conditions. With these tests, it was possible to learn how the specifications of the sensors can affect its performance in different environments. For example, how the baseline of a stereo camera can affect the accuracy depending on the depth.

Check it out in the section Projects!