Las Vegas once again becomes the epicenter of technological innovation with a new edition of Consumer Electronics Show, Where the artificial intelligence and robotics They cease to be a promise of the future and become the central focus of products and services ready to reach the market. More than 4.500 exhibitors and around 150.000 attendees They spend four days touring a fair that has established itself as a great barometer of the trends that will mark the year.
This time, the focus is on how AI is integrated into everyday devices, humanoid robots, and autonomous systemsFrom living rooms to factories and highways, major tech companies and a thriving network of startups are showcasing solutions that combine software capable of reasoning about the physical world with new hardware designs, more precise sensors, and dedicated computing platforms for robots and vehicles.
A fair dominated by integrated artificial intelligence
This year's CES takes place in a global context marked by trade tensions, tariffs and component supply problemsespecially advanced memory and chips, which continue to influence the production of computers, networking equipment, and consumer electronics. Despite this, the aisles of the Las Vegas convention centers showcase a display of new products where the AI appears as an omnipresent claim in posters, keynotes, and demonstrations.
From day one, it's clear that artificial intelligence has ceased to be presented as an isolated function and has become basic infrastructure integrated into processors, sensors, and operating systemsManufacturers are betting on models that run directly on the device, with Greater privacy, lower latency, and the ability to work offline, something crucial for robots, connected cars, wearables and smart appliances.
The big Asian brands —Lenovo, Acer, Samsung, LG or PanasonicCompanies, among others, take advantage of the event to update their laptops, televisions, smartphones, home appliances, and connected car solutions, almost always with some layer of AI as a key selling point. Meanwhile, hundreds of startups are focusing on areas such as... service robotics, digital health, and the smart hometrying to carve out a niche for themselves alongside the giants of the sector.
The fair's schedule maintains the structure of previous years: Preliminary days dedicated to the press and major keynotesThis is followed by four days of open exhibitions combining product announcements, technical demonstrations, and business model sessions. For Europe and Spain, many of the solutions presented serve as a preview of launches that will be arriving throughout the year to local distributors and operators.
Beyond marketing, analysts and innovation managers highlight a change in focus: AI is no longer just presented as futuristic conceptbut as a tool that begins to fit into daily routinesfrom managing home climate control to coordinating fleets of robots in a warehouse.
Nvidia is driving the advancement of AI-based robotics in physics
Among the announcements with the greatest strategic impact, the new offensive of Nvidia in robotics. After consolidating its position in data centers with architectures such as Blackwell and its successor Vera RubinThe company is taking advantage of CES to present a third layer of its offering: Physical AI applied to industrial, humanoid, and mobile robotsThe message is clear: robotics is ceasing to be a limited experiment and is emerging as maturing market.
The stage and associated advertisements feature a succession of demos of humanoid robots, manipulators and mobile platforms developed by partners such as Boston Dynamics, Caterpillar, Franka Robotics, NEURA Robotics, Humanoid, and LG Electronics. The company openly states that the sector is approaching a “ChatGPT-like moment” in robotics, understood as the point where models of perception, simulation and reasoning converge, capable of scaling beyond unique and rigidly programmed tasks.
The key is in the call Physical AIModels that not only analyze text or images, but also They interpret three-dimensional scenes, reason about objects, and plan actions in an environment governed by physical laws. According to the vision presented in Las Vegas, advances in foundational models of the world, high-realism simulators, and vision-language-action architectures They are beginning to unlock applications that until now were economically unfeasible or too fragile for real-world environments.
To implement this strategy, Nvidia is expanding its family of open models Cosmos with variants focused on the generation of physically plausible synthetic dataThe evaluation of control policies in simulation and the understanding of complex scenes using vision and language models. It builds upon this foundation Isaac GR00T, a VLA model designed for humanoid robots that combines contextual reasoning with full body control.
Manufacturers like Franka Robotics, NEURA Robotics or Humanoid They use these workflows to train and validate robotic behavior in simulation before moving to real hardware, reducing costs and risks. This approach is especially relevant for European industrial companies, including those with plants in Spain, where the advanced automation and safety in mixed human-robot environments are growing priorities.
Massive simulation, open robotics, and edge computing
One of the structural problems of robotics has been the fragmentation of tools and processesDifferent platforms exist for simulating, training, testing, and deploying, with specific configurations for each type of robot. New frameworks are being presented at CES that aim to unify these phases through collaborative simulation environments and cloud-native orchestrationfacilitating the work of distributed teams.
Among them, a testing environment stands out, geared towards the standardized evaluation of robotic policies in simulation, connected to benchmarks recognized by the community. The idea is to offer a “common language” to measure the capabilities and robustness of service robots, humanoids, or industrial arms, reducing the gap between research laboratories and production lines.
In parallel, an orchestration system designed for move workloads between local stations and hybrid clouds It attempts to simplify the entire cycle: synthetic data generation, training, software-in-the-loop testing, and deployment. Large cloud providers, such as Microsoft with AzureThey already integrate these flows into their initiatives to accelerate robotics projects.
Nvidia is also reinforcing its commitment to the open robotics through the integration of its models and libraries into hugging face, one of the most active AI ecosystems. VLA models, simulation tools, and components of the Isaac platform are now available for independent developers, universities and R&D centersThis is expected to boost projects in Europe, where research in collaborative and humanoid robotics is particularly intense.
To make all this software work in the physical world, the manufacturer introduces new modules such as Jetson Thor y Jetson T4000, aimed at providing high-performance computing at the edge to robots and autonomous machines. Thor focuses on humanoids and complex robots, while T4000, based on Blackwell architecture, offers an upgrade path for those already using Jetson Orin, with more performance while maintaining limited power consumption.
Humanoid robots and "physical AI" in the home
Beyond the industry, the home service robotics It's grabbing a good portion of the headlines. This year's CES is full of demonstrations of robots that are no longer limited to vacuuming or surveillance, but are starting to grasp and manipulate objects with a certain degree of accuracy, opening the door to domestic applications that recently seemed unattainable.
Among the most visible projects is LG CLOIDa domestic robot presented as household manager capable of performing tasks ranging from starting the laundry to folding and stacking clothes once dry. Unlike other assistants that remain at the digital level, this model incorporates a adjustable torso and two articulated arms with five-fingered hands, designed to interact with everyday objects such as milk cartons or pastries.
The robot moves on a wheeled base and autonomous drivingDesigned to move stably among furniture, people, children, or pets. Thanks to a set of cameras, sensors, and vision algorithms, it is capable of recognize rooms, furniture and utensils, something essential for navigating real homes, not just laboratory environments.
One of the key elements is its integration with the LG ThinQ connected home ecosystemwhich allows it to coordinate other smart home appliances. In demonstrations held in Las Vegas, CLOiD not only manipulated lightweight garments, but also showed how it can Heat a croissant in the oven and bring a jug of milk from the refrigeratorHowever, limitations are still noticeable in more delicate actions, such as pouring liquid into a glass without spilling.
At the same time, LG introduces the brand LG AXIUMa family of robotic components and joints which aspires to become the basis for future models and third-party solutions. The stated goal is to approximate the concept of "home without manual labor", in which routine household chores are taken over by robots and appliances coordinated by AI, freeing up people's time for other activities.
Boston Dynamics, robotic pets and service robots
The CES stage also serves to show how advances in language models translate to robots already knownBoston Dynamics announces the integration of Gemini, Google DeepMind's AI model, in its humanoid Atlas and in the quadruped robot Spot, with the aim of improving the understanding of commands in natural language and interaction with people.
The company is aiming for a gradual transition from pre-programmed movement sequences towards more flexible behaviors, where the robot can interpret complex instructions, adapt to changes in the environment, and manipulate objects with greater autonomy. Emphasis is placed on the long-term goal of achieving robots useful in homes and workplacescapable of handling repetitive or physically demanding tasks, something that is also of interest to European countries with an aging population and a need for care support.
The catalog of robotic proposals is completed with solutions from Mobile robotics for logistics, waiter robots, autonomous inspection systems and devices designed as digital pets equipped with AI, which simulate emotional behaviors and learn from interaction with their owners. Among them, one project stands out: AI-powered pet that physically grows, conceived more as an interactive company than as a functional tool.
Alongside these spectacular demonstrations, some manufacturers are showcasing service robots with a more practical focus: machines capable of restocking shelves, moving loads in warehouses, assisting operators or to conduct routine inspections at industrial facilities. Although many of these solutions are still in the pilot phase, testing is already underway in factories and logistics centers in the United States, Europe, and Asia.
In this environment, the question that looms over many professional sessions is not so much whether the technology will work, but how it will be integrated into real-world environmentsWhat regulations will be necessary and how will the value be distributed among hardware providers, software developers and user companies?
Automotive, robotaxis and software-defined vehicles
La smart mobility It's once again one of the central themes of CES. It's not a typical auto show, but the stands dedicated to electric vehicles, driver assistance systems, and autonomous transportation services occupy a prominent space. The focus has clearly shifted towards the software-defined vehiclewhere remote updates and AI models are as important as the engine or the battery.
In this edition, manufacturers are showcasing cars with more sophisticated AI assistants integrated into the dashboard, capable of managing climate control, entertainment, and navigation using natural language. In parallel, a Partial return to physical check-ups (buttons, dials) in response to the saturation of touchscreens, trying to improve safety and reduce distractions.
Among the most striking announcements is the presentation of a Robotaxi design developed by Uber in collaboration with Lucid and NuroBased on an electric SUV, the vehicle integrates cameras, lidar, and radar, along with a halo light that enhances sensor visibility and serves as an exterior display. The interior is configured to carry up to six passengers with luggage, featuring a large central screen for controlling basic functions and accessing trip information.
The plans envision these robotaxis entering commercial service from 2026Initially in certain cities in the United States. Although the timeline may vary depending on local regulations, these types of projects lay the groundwork for future deployments in other regions, including Europe, where autonomous transportation services are already being tested in limited environments.
In parallel, the automotive sector is exploring the integration of specific AI platforms for reasoning about complex traffic situationsVision-language-action models designed to help automated driving systems manage infrequent or unpredictable scenarios are presented, with pilot tests announced by premium European manufacturers who plan to incorporate these functions into their fleets in the coming years.
Smart home and connected domestic robotics
The segment of smart home It is moving towards systems capable of making contextual decisions, moving away from a simple sequence of programmed commands. Previous leaks and demonstrations already pointed to this. homes that adjust lighting, climate control, audio, security and content depending on the presence of people, the time of day, or learned habits.
Manufacturers are betting on hubs and unified operating systems that coordinate light bulbs, blinds, cameras, motion sensors, and appliances from different brands. The AI analyzes usage patterns, detects anomalies, and suggests routines, so the user doesn't have to configure every detail manually.
In this environment, domestic robotics is integrated as another element of the ecosystem. In addition to ambitious projects like CLOiD, we see vacuum cleaners with robotic armsrobots that move between rooms to Keep an eye on pets, check doors and windows or act as a small mobile security center. Intrusion detection and motion analysis in the home increasingly rely on algorithms capable of interpreting real-time behavior patterns.
Cameras and sensors that use AI are presented for analyze movements and presence Without relying solely on traditional imaging, these solutions integrate techniques that even leverage WiFi signals to map activity in the home. Although these solutions raise privacy concerns, manufacturers insist on processing most data locally, precisely to reduce the need to send sensitive information to the cloud.
For the European market, where energy efficiency and data protection are priorities, these types of proposals are particularly relevant. Systems that optimize electricity consumption by coordinating domestic batteries, solar panels and electric vehicles AI appears as a field of growth, aligned with the European Union's energy transition objectives.
High-performance computing and AI on the PC and TV
CES maintains its role as a showcase for PC hardware and componentsMajor chip manufacturers are introducing new generations of processors for laptops and desktops, with a clear focus: improve local execution of AI models without constantly relying on the cloud. Thinner laptops are expected, with longer battery life and inference engines integrated into the CPU or GPU itself.
The graphics card and AI accelerator sector continues to be constrained by the rising cost of memoirsHowever, incremental improvements in performance and efficiency are being announced. Companies like AMD and Nvidia are detailing plans for new architectures that target more complex workloads, from training medium-sized models to massive inference in data centers and edge computing solutions.
Meanwhile, the universe of screens and televisions It's undergoing its own evolution. Manufacturers are competing to offer brighter panels, with improvements in technologies like multi-layer OLED and advanced backlighting systems that use red, green, and blue LEDs to expand the color gamut. Models are being announced with very high refresh ratesdesigned for video games and high-speed content, and integrate AI functions to dynamically adjust the image.
Artificial intelligence is also making its way into televisions, such as user conversation interfaceSome models include integrated chatbots that allow users to search for content, adjust settings, or receive recommendations using voice and natural language, without needing external devices. For European households accustomed to combining linear television and streaming platforms, these features could facilitate access to increasingly extensive catalogs.
Beyond the living room, unique solutions are presented such as holographic desktop assistants Focused on productivity and gaming, these devices use AI to perform tasks such as programming, translation, document analysis, or simply serving as a virtual desktop companion, anticipating new forms of interaction between people and machines in the workplace and educational environment.
Digital health, wearables and sustainable hardware
La digital health and w They solidify their role at CES as a key category. The new devices go beyond step counting or simple physical activity tracking: they focus on continuous monitoring and early risk detectionAnalyzing sleep patterns, stress, pulse variability, or daily behavior to detect subtle signs of potential problems.
Smart bracelets, watches, and patches that use AI are presented for interpreting large volumes of biometric data and generate personalized alerts, often designed to be shared with healthcare professionals. Manufacturers insist that these devices do not replace doctors, but can complement traditional care, which is relevant for European healthcare systems under increasing pressure.
In a transversal way, the sustainability It becomes a design criterion. Many exhibitors emphasize this. more durable, recyclable, or modular hardwareas well as in low-power AI architectures. There is talk of extending the lifespan of equipment through software updates and the replacement of specific components, in line with European initiatives such as the "right to repair."
The presence of major industrial and energy groups in the conference program underscores the importance of these issues. Topics such as the following are addressed: energy impact of data centers, the long-term opportunities of nuclear fusion or the need for more efficient network infrastructures to support the growth of traffic associated with AI.
In the professional sphere, the sessions on responsible application of artificial intelligence These meetings bring together executives from technology companies and regulatory experts to discuss frameworks for ensuring transparency, security, and data protection. For the European Union, which is advancing its own AI legislation, these discussions are particularly relevant in light of future deployments.
With Las Vegas as a showcase and the AI and robotics As a common thread, this year's CES makes it clear that the next technological wave will not be limited to mobile or computer screens. From robots that fold clothes to autonomous vehicles, from self-adjusting homes to chips designed to reason about the physical world, the fair presents a map in which Europe and Spain must find their own role, combining adoption, regulation, and their own industrial development to seize the emerging opportunities.