Robotaxis in Madrid: when technology takes the wheel

A few years ago, stepping into a driverless taxi would have seemed like something straight out of a science fiction movie. Today, robotaxis have moved beyond the realm of prototypes and experimentation and are edging ever closer to the mainstream. People in cities such as San Francisco, Los Angeles, and Beijing can already ride in self-driving vehicles. Meanwhile, Madrid has given the green light for its first robotaxi trials, opening the door to a new era in which vehicles must do more than simply take us from A to B. They need to interpret their surroundings, make safe decisions, and interact reliably with passengers, infrastructure, and other road users.

For Madrid, the arrival of robotaxis marks a significant step forward in urban mobility. The project will begin with a validation phase that will see 20 vehicles drive around different parts of the city, each with a human safety supervisor onboard. Before they can carry passengers autonomously, they’ll have to pass the real test: dealing with the everyday unpredictability of city streets, from pedestrians and roadwork to emergency vehicles, changing weather conditions, and unexpected situations.

For robotaxis to become a viable transportation option, they must be able to handle all this variability accurately and dependably. And most of all, they must be safe.

Precise positioning: the starting point

A self-driving vehicle needs to know exactly where it is to make safe decisions. This isn’t just another feature; it’s a fundamental requirement. Accurate, reliable positioning allows it to stay in the correct lane, approach stops, comply with restricted zones, and react to obstacles.

Global navigation satellite system (GNSS) technologies are vital in this regard, but getting them to work in urban environments isn’t always so straightforward. Buildings can block satellite signals or cause them to bounce off surrounding surfaces, and self-driving vehicles may need additional information to keep track of exactly where they are. Because of this, they must pull from multiple sources of information and determine not only the accuracy of the data, but also their integrity.

GMV GSharp^® offers a comprehensive, secure, high-precision GNSS positioning solution. By drawing on multiple satellite constellations and frequencies and combining data from inertial measurement units (IMUs), wheel sensors, and other sensing technologies, it can pinpoint a vehicle’s location even in complex environments. Plus, its compatibility with vehicle-to-everything (V2X) applications means the vehicle can put its positioning capabilities to work within a connected ecosystem.

A robotaxi must be able to trust its positioning data when deciding whether to proceed, slow down, stop, or request a safety intervention if its location is uncertain.

From cars to software-defined vehicles

Robotaxis also mark a shift toward the software-defined vehicle (SDV), a model in which many of a vehicle’s capabilities are determined by software and can evolve over time. This makes it possible to introduce new features, improve existing ones and roll out updates remotely throughout the vehicle’s lifespan.

This flexibility is particularly important for robotaxis. Each vehicle is part of a connected fleet, and its software must keep evolving to meet new demands and adapt to changing conditions. But updating a vehicle that makes its own driving decisions is a very different matter from updating a smartphone app. Every change must be carefully tested and validated to ensure that software improvements never come at the expense of safety.

V2X: A car that talks to its surroundings

Autonomous vehicles are better drivers when they can combine what they see through their onboard sensors with information from the world around them. V2X communications make this possible, allowing vehicles to exchange information with infrastructure and other road users.

A robotaxi might detect roadwork using its own sensors, but it could also receive an alert from roadside infrastructure about a lane closure or an approaching emergency vehicle. This shared information helps the vehicle anticipate what’s ahead, making driving smoother and safer. It can also help vehicles move through regulated urban zones, manage priority at certain points, and coordinate with traffic control centers.

GMV has experience in cooperative intelligent transportation systems (C-ITS) and V2X communications, including ITS-G5 and cellular 5G C-V2X technologies, as well as connected mobility projects on roads, highways and in urban environments. This expertise helps position the robotaxi not as a standalone vehicle, but as part of a wider intelligent mobility network.

Cybersecurity by design

A robotaxi is, after all, a huge, complex connected system on wheels, bringing together software, connectivity, location data, and cloud services. And the more connected it is, the more important it becomes to keep it secure.

Protecting a connected, self-driving vehicle means preventing unauthorized access, identifying vulnerabilities, responding to incidents, and maintaining constant vigilance throughout the service’s lifecycle. GMV brings expertise in this area, with capabilities aligned with standards such as ISO/SAE 21434 and UNECE R-155, alongside services covering vulnerability management, threat analysis, and security operations for connected products and vehicles.

Cybersecurity isn’t just another technological feature. It’s essential to earning the trust of passengers, operators, and public authorities.

In-cabin intelligence: safety starts on the inside

Taking the driver out of the equation introduces another interesting challenge: someone still needs to know what’s going on inside the vehicle.

Has everyone gotten in? Has a child been left behind after a ride? Has someone forgotten an item? Is something happening that requires assistance?

In-cabin technologies use cameras, radar, and intelligent data processing to understand what’s going on inside the vehicle. And in a robotaxi, where there’s no driver to keep an eye on things, these technologies take on a whole new significance. At GMV, we draw on our expertise in computer vision, artificial intelligence, and safety-critical software development to design and integrate systems that detect and classify occupants, objects, and unusual events, helping make journeys safer and more comfortable. When used effectively, these technologies can help robotaxis respond to passengers’ real needs, whether that’s checking that nobody has been left inside, raising an alert when something unusual happens, or supporting remote assistance.

Of course, all these capabilities must be developed with strict safeguards for privacy and data protection.

Conclusion: turning autonomy into trust

The arrival of robotaxis in Madrid presents an opportunity to make urban mobility more accessible, efficient, and sustainable. But their success won’t depend simply on whether a vehicle can drive itself. What matters is whether it can do so safely and consistently while sharing the city’s streets with everyone else.

Precise, secure positioning, robust SDV architectures, V2X connectivity, cybersecurity, and onboard intelligence are all pieces of the same puzzle. Autonomous mobility will become a reality when these technologies work seamlessly together to make every journey one passengers can trust.

So, when we hail a taxi a few years from now and find the driver’s seat empty, perhaps we won’t be asking who’s driving anymore. The more interesting question will be how many technologies are working together behind the wheel.

Autor: Beatriz García