AI Robustness & Trustworthiness Engineer – Computer Vision for Safety-Critical Aerospace Systems
Alicante,
Spain
#Join us in enabling drones to populate our skies
At Embention, we have been driven since 2007 by a clear mission: Enabling drones to populate our skies.
We provide more than 700 customers in 70 countries including internationally renowned companies such as Toyota, Amazon Prime Air, and Honda with Veronte autopilots, critical avionics, and software for their professional drones and eVTOLs. All our systems are developed in accordance with DO-178C, DO-254, and DO-160 standards, making us the only company in the sector certified by AESA and EASA as both a POA and APDOA.
The core of our company lies in our R&D team, made up of over 100 multidisciplinary engineers. Together with the rest of the organization, we are a team of 180 professionals united by a single goal: to innovate and push the boundaries of UAV technology.
Following a record-breaking 2025 marked by strong stock market performance, the launch of industrial production in the United Arab Emirates, and the establishment of a subsidiary in the United States, we are looking to continue expanding our team and driving continuous R&D development.
If you are passionate about our mission and want to be part of this journey, we are looking for you.
🌟 What would be your mission?
Your mission will be to ensure the robustness, traceability, and technical defensibility of the Detect and Avoid (DAA) system's computer vision model, guaranteeing its operational reliability in BVLOS environments and its alignment with EASA regulatory frameworks and applicable EUROCAE WG-114 guidelines for machine learning-based systems.
💡 Which will be your responsibilities?
Design and execute the robustness strategy for the vision model.
Evaluate model behaviour on out-of-distribution (OOD) data.
Implement or integrate anomaly detection mechanisms.
Analyse and characterise epistemic and random uncertainty.
Evaluate the model's confidence calibration.
Monitor the quality, coverage, and governance of datasets.
Implement complete data-model-results traceability.
Collaborate with the simulation team to generate adversarial scenarios.
Apply explainability techniques (Grad-CAM, Score-CAM, or similar) to analyse the model's behaviour and detect spurious correlations.
Document operational limitations of the detector and its conditions of validity.
Contribute to technical arguments aligned with regulatory frameworks (EASA AI guidelines, WG-114, ED-324 where applicable).
🕵️♀️ What do we need?
Degree in Computer Engineering, Telecommunications, Mathematics or similar.
Proven experience in training and advanced validation of computer vision models.
Experience in versioning datasets and models (MLOps).
Data quality and governance applied to images.
📚 What do we value?
Knowledge of out-of-distribution, anomaly detection, explainability.
Ability to translate technical results into structured and defensible documentation.
Previous experience in aeronautical or safety systems.
Knowledge of SORA, ARC or UAS regulatory frameworks.
Familiarity with EASA and EUROCAE WG-114 documents and recommendations.
Experience in regulated or certifiable environments.
Embention is an equal opportunity employer. Recruitment decisions are made based on experience, qualifications, and alignment with role requirements.