For years, the autonomous vehicle industry was consumed by a debate: cameras versus lidar, lidar versus radar, vision-only versus sensor fusion. Few voices were louder than Elon Musk’s. In 2019, Musk famously referred to lidar as a “crutch” and a “fool’s errand”, arguing that camera-based systems alone would ultimately be sufficient for autonomous driving. 
At the time, the comment sparked endless debate across the automotive and technology industries. Supporters of lidar argued that precise 3D measurements were essential for safe autonomy. Vision-only advocates countered that cameras, combined with increasingly powerful AI, could provide all the information a vehicle needed to navigate the world. Years later, however, the conversation has evolved. 
The question is no longer whether vehicles should use lidar, radar, cameras, or some combination of sensors. Instead, the industry is increasingly focused on a more important challenge: how sensing, AI, and computers work together to help machines safely understand and interact with the physical world. The future is not about choosing a single sensor. It’s about building perception systems capable of delivering reliable, real-time understanding of complex environments. 

Self-driving technology is one of the most exciting advancements in the automotive industry. While it promises safer roads and greater convenience, the price of developing and implementing autonomous vehicles (AVs) is incredibly high. From advanced sensors to powerful computing platforms, every part of an AV comes with a hefty price tag. In this article, we will break down the costs of self-driving technology, component by component, so you can understand where the money goes and what this means for the future of transportation.

Advanced end-to-end autonomous driving systems predict other vehicles' motions and plan ego vehicle's trajectory. The world model that can foresee the outcome of the trajectory has been used to evaluate the autonomous driving system. However, existing world models predominantly emphasize the trajectory of the ego vehicle and leave other vehicles uncontrollable. This limitation hinders their ability to realistically simulate the interaction between the ego vehicle and the driving scenario. In this paper, we propose a driving World Model named EOT-WM, unifying Ego-Other vehicle Trajectories in videos for driving simulation. Specifically, it remains a challenge to match multiple trajectories in the BEV space with each vehicle in the video to control the video generation. We first project ego-other vehicle trajectories in the BEV space into the image coordinate for vehicle-trajectory match via pixel positions.

Humans have a remarkable ability to learn to drive quickly and can obtain a licence to drive across a whole country after tens of hours of practice. But after 10 years of commercial self-driving car development, over 10 million autonomous miles and $5B per year spent, we still do not have commercial self-driving vehicles on our roads. To turn what is currently a fantasy into a reality, we need to take a different approach.

Autonomous driving in Europe faces a multifaceted array of challenges, but our review indicates that these challenges are well-understood and, with concerted effort, can be addressed. The implications of these findings are significant. In the short term, the persistence of regulatory uncertainty, technical limitations, and public scepticism means that fully self-driving vehicles will likely roll out slower in Europe than the most bullish predictions once suggested. Stakeholders must navigate a delicate balance between innovation and caution, since premature deployment could erode trust if incidents occur, whereas excessive delay could cause Europe to fall behind in a transformative industry. In the long term, overcoming these barriers is crucial because the potential benefits of AVs - from dramatically reduced traffic accidents and enhanced mobility for the elderly or disabled, to more efficient logistics and reduced congestion - are substantial. Failure to resolve the impediments could mean foregone benefits in safety and economic efficiency, not to mention Europe’s competitiveness in the global automotive arena.

The company announced earlier this month a deal with Dongfeng to build Jeep and Peugeot models for the Chinese market and increase cooperation on technology research and development.
Dongfeng and other Chinese carmakers meanwhile are looking to Europe and other export markets as their home market remains tough, with consumer spending slumping.
Brands such as BYD, Chery, Geely, Leapmotor, Jaecoo, and XPeng were virtually unknown three years ago in Europe.

The Bosch Center for Artificial Intelligence (BCAI) serves as Bosch’s center of excellence, dedicated to integrating cutting-edge AI technologies into the company's products and services to create solutions truly “Invented for life.” By leveraging Bosch's vast domain expertise and world-class AI methods, BCAI actively spearheads applied AI projects from initial concept to implementation, bridging the gap between fundamental research and real-world applications. With a strong focus on skills, agility, and openness, BCAI teams, comprising domain experts, data professionals, and software engineers from diverse international backgrounds, are at the forefront of AI research and application.

The 2026 revision of the Current Procedural Terminology (CPT) codes for lower extremity revascularization introduces important changes that are intended to better align code selection with contemporary endovascular practice.1 In this article, I highlight several key updates and illustrate how I applied the revised codes to a case of chronic limb-threatening ischemia (CLTI) in the setting of multilevel occlusive peripheral artery disease (PAD).

We believe intelligence that cannot act in the physical world is incomplete. The next frontier of AI is not only to understand information, but to operate safely and intuitively in the real world. It will be defined by systems that can perceive, reason, learn, and make decisions safely in dynamic physical environments. Systems that understand uncertainty, causality, motion, interaction, and consequence.
That requires a different set of problems to be solved: world models that support action, representations grounded in space and physics, policies that adapt under uncertainty, and learning systems that improve through interaction with reality rather than passive observation alone.
This is the challenge of embodied AI.

Uber et la start-up Pony.ai lancent des essais de voitures autonomes destinées au marché européen. Ce partenariat vise à explorer le potentiel des robotaxis sur le continent, une étape stratégique pour l’essor de la mobilité sans conducteur en Europe.

La singularité technologique de Wayve réside dans sa pile logicielle « AV2.0 », qui s’appuie sur un modèle d’IA généralisée end-to-end. Contrairement aux approches classiques (architecture « sense-plan-act »), l’algorithme de conduite autonome traite directement les données brutes des capteurs (caméras, radars, voire LiDAR) pour piloter le véhicule. Concrètement, cela permet à un robotaxi équipé du système Wayve d’apprendre à circuler dans des environnements urbains variés sans recourir systématiquement à des cartes HD ou à une phase d’entraînement longue et locale.

Stellantis et Microsoft annoncent une collaboration stratégique de cinq ans visant à accélérer la transformation numérique du groupe automobile. Co-développement de plus de 100 initiatives d’intelligence artificielle, renforcement de la cybersécurité, modernisation de l’infrastructure cloud avec un objectif de réduction de 60 % de l’empreinte des data centers d’ici 2029 : l’accord couvre l’ensemble de la chaîne de valeur du groupe, de l’ingénierie à l’expérience client.

 Nissan has unveiled its ambitious future product strategy, as detailed by automotive journalist Alisa Priddle, on April 14, 2026. This plan highlights a transformative approach for the company, focusing on enhancing its global  vehicle lineup with a greater emphasis on hybrid technology, reducing the number of models offered, and integrating advanced artificial intelligence (AI) solutions into its operations. These strategic shifts are designed to position Nissan favorably amid the evolving landscape of the automotive sector.

Trajectory prediction is critical for autonomous vehicles, allowing them to anticipate the future movements of other agents on the road to ensure safety and smooth navigation. Traditional methods operate under the assumption of zero latency—an idealized scenario where predictions are made instantaneously without delay. However, in practical autonomous driving systems, latency is unavoidable due to the time required for data sensing, processing, and prediction calculations. The conventional oversight of this latency leads to predictions that are already outdated the moment they are generated, resulting in reduced accuracy and potential safety risks.

Le 15 avril 2026, Le Journal de l’Automobile a remis le titre de Personnalité Automobile de l’Année 2025 au directeur général de Valeo. Cette cérémonie a rassemblé les grands acteurs de la filière à la Maison de la Chimie, à Paris, pour saluer le rôle central de Christophe Périllat dans la transformation de l’industrie européenne et son engagement en faveur d’une politique industrielle ambitieuse.

A year ago, MOIA America and Uber unveiled an ambitious plan to launch a commercial robotaxi service — using autonomous versions of Volkswagen’s electric ID. Buzz minivan — in multiple U.S. cities over the next decade. Los Angeles is the first city on that list.
Testing will start in the next few weeks with about 10 autonomous ID. Buzz vehicles in Los Angeles, according to the company. The production version of the driverless vehicle seats four people.

While humans are responsible for killing animals with their cars all the time, this incident has brought negative attention to the new technology. Local media picked up on the duck incident after a resident posted in a Mueller neighborhood Facebook group that an Avride autonomous vehicle (with a human safety operator behind the wheel) ran over and killed a duck, and did not stop afterwards. “It didn’t slow down or hesitate at all, just steamrolled through,” the post, which KXAN reported on, reads.