3 juin 2024 | International, C4ISR
AI regulators fear getting drowned out by hype of wars
Advocates for limiting artificial intelligence in combat say the global security environment isn't conducive to their cause, but interest remains.
10 février 2024 | International, Terrestre
3 juin 2024 | International, C4ISR
Advocates for limiting artificial intelligence in combat say the global security environment isn't conducive to their cause, but interest remains.
11 février 2021 | International, Aérospatial
Unveiled with pomp at Aero India 2021, the largest airshow since the start of the pandemic, the HAL Combat Air Teaming System (CATS) looks a bit derivative, with its centerpiece – the CATS Warrior – looking almost identical to the Kratos Valkyrie, a drone that captured the imagination of aviation community several years ago. The resemblance is not coincidental. Drones of this kind are informally called “loyal wingmen”, and they are often compared to unmanned fighter jets. Currently under development with most leading military powers, they are set to be controlled by artificial intelligence (AI) instead of ground-based operators, and accompany manned fighter jets into battle. In the United States, the Skyborg program is aimed at developing loyal wingmen for the US Air Force. In Europe, the Mosquito will soon be flying with the Royal Air Force (RAF), while the Future Combat Air System (FCAS) has at least several designs in the works. Russia has been working with the concept too, as did China and some other countries. An ambitious project On paper, theCATS looks very similar to all of those developments. According to the Indian press, it is going to be comprised of several interconnected systems: First off, the whole idea revolves around “Mothership for Air teaming eXploitation” (MAX), a modified two-seater variant of the HAL Tejas Mk-1A fighter jet designed to control a number of drones in flight. It would carry the CATS Hunter, which is described as a fighter-launched cruise missile that would have a range of 700 kilometers (435 miles) with a regular warhead. In a different configuration, the Hunter would have a range of 350 kilometers (217 miles) and could return to base for reuse. Its payload, then, would consist of Air Launched Flexible Assets (ALFAs), swarming munitions each carrying 5 to 8 kilograms of explosives and likely similar in its concept to loitering munitions used by many modern armies. A mockup displayed at Aero India 2021 showed four ALFAs in an internal cargo bay of one Hunter. The last component of the CATS program would be the Warrior drone, a loyal wingman with stealth features, powered by the domestically-produced PTAE-7 turbofan engine and carrying a pair of air-to-air missiles, ALFAs or laser-guided bombs in its two internal bays. With an active electronically scanned array (EASA) radar, Electro-Optical/Infra-Red (EO/IR) imaging system and electronic warfare suite, it could be used both as a forward-deployed scout for regular aircraft as well as for directly engaging enemy targets. It is important to understand that so far these projects are in a development stage. HAL claims that it has been working on the concept since early 2018, but the development really started only in late 2019 and early 2020. The deadline is scheduled for 2024-2025, which could seem optimistic for regular aircraft, but falls in line with similar projects: both the Skyborg and the Mosquito aim at initial operational capability by 2023. Crucial differences There are several key differences between the CATS and other similar programs though. First off, the CATS Warrior is the first loyal wingman showcased, at least in mockup form, with air-to-air missiles. Many manufacturers of prospective loyal wingmen have hinted at such a capability, yet they tend to be careful with their claims. The reason for that is clear: while it is relatively easy to make a drone capable of launching infrared-guided missiles, the participation in actual aerial combat, especially if such a drone is partially or primarily AI-controlled, is a whole other level of complexity. It is very likely that the first “generation” of loyal wingmen will have only rudimentary air-to-air capability and the option to engage in a pitched aerial combat will come later, with upgrades, refinements or subsequent programs (such as the DARPA's LongShot). Both Kratos and Boeing, two companies that already developed and tested their loyal wingmen, talk quite assertively about reconnaissance and ground attack capabilities of their aircraft, but hint at air-to-air capabilities with far less certainty. The two aforementioned drones are supposed to be modular though, their components, such as detection or payload delivery systems, being mission-adaptable. The modularity of the CATS Warrior was not mentioned by HAL at the airshow, and the existence of the multi-purpose Hunter is partially compensating for its lack. Yet another large difference between the CATS and rival Western programs is an emphasis on AI control. It is quite clear that although ALFAs will likely use some form of artificial intelligence, the existence of dedicated two-seater control aircraft hints at Warrior being, at least in some part, piloted. According to HAL, its loyal wingman will be capable of autonomous take-off and landing, yet the capability of autonomous combat was not revealed – an element which, if planned, would likely become its main selling point. In this regard, India is not alone, as the Russian Grom is intended to be human-controlled too, at least according to the current plan. But both American and European programs dedicate a lot of effort and investments into the development of AI capable not only of controlling swarms of combat drones, but of taking over part of the pilot's workload too. Human-AI teaming proved to be a difficult concept, necessitating the development of special algorithms and interfaces with features not explored before. Reacting to circumstances That might be the reason HAL keeps conservative with the control possibilities of its loyal wingman. The ground has not been proven for autonomous fighter jets, and being a pioneer in this field requires colossal research and development funding – money that would be better spent on more pressing issues. Such as the lack of fighter jets. India has been struggling with that for some time now, introducing a hotchpotch of models – from brand new Dassault Rafales to refurbished 80s-vintage MiG-29s – just to close the air defense gap. The ramping up of the production of the HAL Tejas was not enough for that too. The latest MRCA (multi-role combat aircraft) competition has been dragging for some time now, and even if India finally selects its new fighter jet, it will take quite some time to reach operational capability. The CATS Warrior can be interpreted as a direct response to that. If the whole project will enter mass production by the mid-20s, as expected, it may become an ultimate way to solve IAF's long-running problem without greatly increasing the production of the Tejas. With an advertised cost of $5 million per unit – more than most Western loyal wingmen, but still negligible in comparison with manned jets – it could be a saving grace for the country. https://www.aerotime.aero/27216-What-do-we-know-about-Indias-fighter-jet-drone-program
23 janvier 2020 | International, Aérospatial
EXCLUSIF Pierre-Eric Pommellet, directeur général de Thales, a été choisi par l'APE pour succéder à Hervé Guillou à la direction de Naval Group. Son nom doit encore être avalisé par l'Elysée. D'autres noms ont circulé, comme celui du directeur général adjoint de Naval Group Alain Guillou, celui du directeur des programmes Olivier de la Bourdonnaye ou celui de Marie-Pierre de Bailliencourt, ancienne DG du groupe, qui avait été un des grands artisans du contrat du siècle des sous-marins australiens. Le nom de Benoît Ribadeau-Dumas, directeur de cabinet d'Edouard Philippe et ancien de Thales et Zodiac, avait aussi été cité, au grand agacement de l'intéressé. Le scénario d'une prolongation d'Hervé Guillou a aussi été évoqué : celui-ci étant touché par la limité d''ge de 65 ans le 24 mars prochain, un tel projet nécessiterait un changement des statuts du groupe. Pierre-Eric Pommellet est tout sauf un inconnu pour le petit milieu de la défense. Né à Brest, ce polytechnicien affable, surnommé « PEP », a débuté à la Direction générale de l'armement en 1990, avant de passer deux ans à la DCN (Direction des constructions navales), l'ancêtre de Naval Group. Un passage en cabinet ministériel plus tard, chez Jean-Pierre Raffarin au ministère de l'artisanat, du commerce et des PME, il entre chez Thales dont il gravit peu à peu les échelons : directeur de l'usine du Haillan (Gironde), directeur des équipements militaires, directeur de la division aérospatiale, puis de celle des systèmes de mission de défense. Lors du départ du patron de Thales Jean-Bernard Lévy vers EDF fin 2014, Pierre-Eric Pommellet est un des favoris pour le fauteuil de PDG. Il peut même se prévaloir du soutien de Dassault Aviation, actionnaire à 25% du groupe. Mais le conseil lui préfère finalement Patrice Caine. Cette déception n'empêchera pas les deux hommes de bien travailler ensemble : Patrice Caine nommera même Pierre-Eric Pommellet directeur général en charge des opérations, en clair numéro deux du groupe. « L'entente entre les deux hommes est très bonne », assure un connaisseur de la maison Thales. https://www.challenges.fr/entreprise/defense/le-numero-deux-de-thales-favori-pour-diriger-naval-group_695054