Safran Electronics & Defense has expanded its simulation portfolio with the release of Skydel AI, a tool designed to streamline GNSS simulation automation for engineers, integrators, and researchers working on advanced positioning, navigation, and timing (PNT) applications.

The new solution addresses a common challenge in GNSS testing: configuring complex simulation environments that replicate real-world signal conditions. Traditionally, these setups required extensive scripting, deep domain expertise, and iterative testing. Skydel AI introduces a natural language interface that translates user commands directly into Python code, significantly reducing time-to-test while improving repeatability.

Technical Capabilities

At its core, Skydel AI integrates with Safran’s proven Skydel simulation engine. Engineers can request modifications in plain English—such as creating multipath conditions, simulating jamming scenarios, or adjusting orbital parameters—and the system generates the corresponding Python scripts automatically. This reduces manual coding errors and accelerates prototyping cycles.

---

Also Read – Trimble Tackles GNSS Spoofing with First Multi-Constellation Anti-Spoofing Tech

---

One of the standout innovations is an AI-driven tropospheric delay model. Built on a dataset of 14 million samples from 221 GNSS reference stations, the neural network processes real-time atmospheric data sourced via the Open-Meteo API. This architecture enhances wet delay modeling accuracy by up to 88% compared with legacy empirical models, enabling more realistic replication of weather-dependent signal disturbances.

Engineering Applications

For system developers and researchers, Skydel AI offers flexibility in designing repeatable, high-fidelity scenarios for testing multi-constellation and multi-frequency receivers. By automating scenario generation, teams can reduce test cycle times and focus on evaluating receiver performance under controlled but realistic conditions.

Safran also emphasizes backward compatibility with existing Python workflows, allowing engineers to incorporate Skydel AI into established test pipelines without overhauling infrastructure. This integration ensures that teams can leverage existing automation scripts while benefiting from natural language inputs.

Industry Relevance

According to Pierre-Marie Leveel, Safran’s Program Director for PNT, the development is aimed at shortening lengthy test processes while maintaining simulation accuracy. Given the increasing complexity of GNSS environments—driven by multi-constellation signals, interference sources, and atmospheric variability—tools like Skydel AI represent a step toward scalable GNSS simulation automation in both research and operational test facilities.

Safran’s legacy in PNT simulation provides a foundation of trust and reliability. With Skydel AI, the company demonstrates a continued focus on delivering robust, flexible, and future-proof tools for the GNSS engineering community.

Source: SAFRAN