Rafale Jet to Detect Targets 100 Times Faster with New AI-Infused Talios Pod

Thales has upgraded its Talios targeting pod with artificial intelligence that will enable the Rafale fighter jet to detect targets 100 times faster than the current human-dependent system by automating imagery analysis.

Since it entered service in 2018, the Talios reconnaissance and targeting pod has steadily added new functions to enhance operational value. And starting with the Rafale standard F4.3 in 2026, its deep learning algorithms will be capable of searching for objects of interest in a given zone 100 times faster than the current system that involves sending the captured imagery to ground control which in turn provides target data to the pilot.

Operational Deployment by 2026

Thales first presented the potential of the AI to the French armed forces in 2018, and a contract was awarded in December 2023. The new AI function is expected to enter service in 2026 with the arrival of the Rafale standard F4.3, and will be the first function on board the Rafale to make such intensive use of deep learning technologies.

This latest innovation is one of the most important developments so far by cortAIx, the artificial intelligence accelerator set up by Thales to leverage the Group's extensive AI expertise and build ground-breaking solutions for the armed forces, aircraft manufacturers and other operators of critical systems, a Thales release said.

On-board image analysis in real time

The AI installed in the pod  scans the combat zone, automatically analyzing the images captured and telling the pilot what it's detected. By pre-selecting objects of interest, it reduces the cognitive load on the pilot, but the decision to engage a target remains the pilot's responsibility at all times.

Importantly, because the AI is installed in the pod itself – despite physical challenges related to temperature, vibration and energy consumption – it provides information in real time as the mission unfolds.

This overcomes the need for a datalink to send imagery to a ground station, because everything happens on board the aircraft. In addition, this AI is capable of spotting small objects in the images, enabling the pilot to remain at a safe stand-off distance from any potential threats. All of which speeds up the tempo of operations because the pilot is kept informed about the tactical situation in real time.

Co-engineered with operational personnel

The AI developed for the Talios pod draws on several years of R&D and was trained on vast numbers of examples sourced from sovereign imagery databases compiled by Thales during test flights or provided by the French armed forces. In addition, thanks to an in-house innovation lab called Image'Inn, operational personnel were placed in realistic situations to test different scenarios and fine-tune the user interface. Thanks to this co-engineering approach with end users, the new function has been developed well ahead of operational deployment.

Towards collaborative combat

The longer-term significance of the AI installed in the pod will be felt in the era of collaborative combat, which will rely on data exchanges between the different assets deployed in the theatre as part of sixth generation fighter project. Given the huge volumes of data generated by all the sensors, the AI will play an essential role by only extracting and transmitting relevant information, overcoming the danger of saturating communication systems.