The dream of fully autonomous driving seems to be always just around the corner, but the reality is that there are still significant technological challenges to overcome. Most of the companies leading the development of this technology, such as Waymo, are betting on high-definition maps as a fundamental pillar of their systems. These maps provide accurate information about the environment, facilitating vehicle navigation and decision-making. However, a new player in the field, Imagry, proposes a radically different approach: autonomous driving without map support.
The company's website can be viewed at: https://imagry.co/
Imagry, a company based in California and Israel, is developing an autonomous driving system that relies solely on the perception of the environment through sensors, similar to Tesla's philosophy. Instead of relying on predefined maps, Imagry's system "learns" to navigate by interpreting in real time the images and data captured by its sensors. This means that the vehicle is able to recognise traffic signs, pedestrians, other vehicles and terrain features without needing prior information about the area.
The company is testing its technology on autonomous buses in Israel, with the ambitious goal of achieving Levels 3 and 4 of autonomy (SAE) without the use of maps. The choice of buses is not accidental. These vehicles operate on pre-set routes and at relatively low speeds, providing a controlled and predictable environment for the development and refinement of the system.
In addition, the buses face a wide variety of traffic situations in urban environments, allowing the system to learn and adapt to different scenarios. Passengers can request the bus to stop by pressing a button or through a mobile application, simulating real-world interactions.
Imagry argues that its approach, based on computer vision, is more flexible and scalable than map-based systems. Creating and maintaining high-definition maps is a costly and complex process, requiring constant updating to reflect changes in the environment. On the other hand, a vision-based system can adapt to new environments without the need for human intervention, simply by “watching” and learning from its surroundings.
This approach, while promising, also presents very significant challenges. The accuracy and reliability of environmental perception are crucial to the safety of the vehicle and its occupants. The system must be able to correctly interpret camera images and sensor data in real time, even in adverse conditions such as rain, snow or low light.
In addition, the ability to anticipate and react to unforeseen situations, such as erratic behavior by other drivers or the presence of unexpected obstacles, is critical to ensuring safe driving.
The similarity with Tesla’s approach is obvious, but Imagry aims to go further. While Tesla's Autopilot, despite its advances, still requires driver supervision and is not considered a fully autonomous system, Imagry is seeking to achieve Level 4 autonomy, where the vehicle can handle most driving situations without human intervention.
One of the most attractive aspects of Imagry's system is its cost. The company says its technology is priced at around $3,000, significantly less than the $10,000 price tag for Tesla's Autopilot.
If Imagry can prove the reliability and effectiveness of its system at that price, it could have a considerable competitive advantage in the market. This could open the door to mass adoption of autonomous driving technology, not only in private vehicles, but also in public transport fleets and commercial vehicles.
Imagry's ultimate goal is to offer its system to car and heavy vehicle manufacturers. The experience gained from testing on buses provides a solid foundation for adapting the technology to other types of vehicles.
A video can be seen at: https://youtu.be/tLo2JyYP2q4
Imagry is still a very small company, and its technology is still in the development phase. The road to fully autonomous driving is long and full of obstacles. However, the company's innovative approach, based on artificial vision and without dependence on maps, represents a promising alternative to traditional systems.
If the company manages to overcome the technical challenges and demonstrate the reliability of its system, it could revolutionize the future of mobility. Time, as always, will tell whether this vision becomes a reality.
EN 
ES