Davide Canavesi testing software for autonomous driving

Davide Canavesi wants to put autonomous driving out into the world, honing LiDAR innovations that are central to safe autonomous driving at a start-up supported by ZEISS. His vision of self-driving cars embodies safer mobility. Navigating innovation, he's an advocate for a transformative era defined by safety and sophistication. Canavesi’s efforts carve out a horizon where autonomous driving is not just a concept, but a constant stream of progress. The tech is coming, and he’s ready to show a future with zero accidents.

At a shooting range on the outskirts of Ulm, Germany, Davide Canavesi and his team are meticulously aligning equipment and sending a laser at a target 300 meters away. Unlike in Hollywood movies, this laser won't cause an explosion. Instead, it's revolutionizing the automotive industry. “It's all about accuracy and control," Canavesi explains, "even at long range.”

Davide Canavesi inspecting an optical sensor

Sensing and steering the future of mobility

Canavesi works as the Head of Strategy and Business Development at Scantinel Photonics, a spin-off company of ZEISS. Scantinel specializes in the development of cutting-edge optical sensing technology. Among their innovations is pioneering work on FMCW LiDAR chips: advanced components used in remote sensing systems that pave the way for full autonomous mobility. FMCW LiDAR stands for “frequency-modulated continuous-wave light detection and ranging,” a technology crucial for various applications such as autonomous vehicles and environmental monitoring.

These sensors enhance driving safety by perceiving objects and anticipating road events up to ten seconds in advance. The significance of this technology is clear to Canavesi. “You need to identify the height of the tire or any obstacle to determine whether to drive over it, detour or brake,” he says.

Traffic remains dangerous, with the European Commission reporting over 20,000 deaths in the EU in 20231, and the US seeing over 40,000 deaths annually. In 94% of cases, human error contributes to the accident.2, 3 Canavesi had his own encounter with a dangerous traffic accident. Twenty years ago, while driving back from school in Italy, another vehicle failed to stop at a stop sign and collided with his car, causing it to flip over. His car slid across the road, and he had to escape through the rear windshield.

Portrait of Davide Canavesi

Humans can only guess distances. We need fully automated systems that can actually measure them. That’s where we come in.

Davide Canavesi

Head of Strategy and Business Development | Scantinel Photonics
Davide Canavesi taking a walk with his daughter

Shifting gears: advocating for automated safety

Despite the harrowing experience, Canavesi explains on his way back to Scantinel's office in Ulm in Germany that he emerged relatively unscathed. Nevertheless, the accident showed him one thing, that "humans can't control cars perfectly.” Therefore, to overcome human limitations, he advocates for self-driving cars enabled by AI and machine learning. "Humans can only guess distances. We need fully automated systems that can measure them. That’s where we come in.”

When Canavesi became a father of two, the issue of safety took on even greater importance. His oldest daughter, now eight, rides her bike to school every day. "I live with my family in Munich. In a city like this, the traffic situation is nerve-wracking,” says Canavesi. He's all too aware that his daughter's safety depends not only on her own skills, but also on the unpredictable behavior of countless drivers. "Essentially, it's a gamble. So, anything that can make that gamble safer should be done,” he says.

And there’s so much potential to improve those odds. The human brain processes an astonishing 11 million sensory impressions per second, yet only forty of these are perceived consciously, leaving room for significant error.4 In contrast, machines, theoretically, have the capability to perceive everything they are programmed to detect. This fundamental difference underscores the potential precision and reliability of autonomous driving compared to the often fallible human sensory experience.

  • 1,350,000

    people die in traffic accidents every year.5

  • Every 24s,

    someone is killed on a road around the world.5

  • 94%

    of deadly road accidents stem (at least in part) from human error.3

  • By 2050,

    the EU wants to reduce the number of road deaths to zero.6

  • Up to 10s

    in advance, LiDAR sensors from Scantinel Photonics detect objects and anticipate events on the road, enabling safer autonomous driving.

Davide Canavesi working on safe autonomous driving

Building trust in autonomous driving

Scantinel Photonics, Europe's only manufacturer of the pioneering FMCW sensors to make this transition possible, has grown from a staff of six to over sixty since 2019 through collective effort, with Davide Canavesi being one of the very first to join the team. Their mission is to equip autonomous cars with the ability to see and react to road conditions far beyond the capabilities of any human driver.

Scantinel's FMCW LiDAR technology is revolutionizing autonomous journeys by overcoming traditional limitations with enhanced resolution and range. This innovation is transforming object detection and environmental perception, even in challenging weather. By integrating the FMCW LiDAR into a silicon chip, the company enables cost-efficient mass production of sleek, self-driving cars without bulky components.

In this mission, Canavesi acts as a bridge between customers and technology, with ZEISS pushing the start-up forward with the necessary resources, ensuring everything runs smoothly. "It's like enabling technology to soar. You don't confine the bird to its nest, you let it fly for the rest of the world,” Canavesi says.

An optical sensor in the lab

Yet the race between innovation and regulation persists – potentially hindering the integration of evolving technology in transportation. New EU regulations for driver assistance systems took effect in July 2024, while the UN plans to put regulations for Level 3 and Level 4 autonomous driving into effect in January 2025. These regulations expand coverage to include features like lane-changing systems on various road types but do not address fully autonomous driving, still prioritizing driver engagement and awareness.

Canavesi says the technology is developing much faster than what the current legislation supports. “Our predictions for the next five to ten years are often off. This is largely because our visions tend to be too conservative. Technological advances can outpace even the most ambitious projections,” he says as he enters the Scantinel building.

Building public trust in driverless cars is another hurdle. Canavesi, passionate about making technology understandable, sees his role as demonstrating the future benefits. He believes that the technology they’re working on in driverless vehicles could reduce traffic accident deaths to zero, just like the EU hopes to do by 2050. “We are struggling with the issue of the moment, but my task is to show the future benefits,” he says as he guides us through the laboratories. He plans on continuing to work with his customers to really answer the question: what makes mobility safer? To him, it’s reducing risk as much as possible and going beyond what a normal road user can sense.

Improving sensors and enabling autonomous driving would give people time to do other things while commuting. Imagine what people could accomplish if they could focus on other things in the car instead of driving.

Davide Canavesi

Head of Strategy and Business Development at Scantinel Photonics
  • Visualization of LiDAR sensor data
  • Autonomous driving level zero
  • Autonomous driving level one
  • Autonomous driving level two
  • Autonomous driving level three
  • Autonomous driving level four
  • Autonomous driving level five
  • Visualization of LiDAR sensor data
  • Visualization of autonomous driving level zero, no assistance
  • Visualization of autonomous driving level one, driver assistance
  • Visualization of autonomous driving level two, parking assistants
  • Visualization of autonomous driving level three, cars taking over depending on traffic situation
  • Visualization of autonomous driving level four, vehicle takes over all driving functions
  • Visualization of autonomous driving level five, vehicle can operate without driver

In motion: a look ahead

After showing the Scantinel facility, the rest of the day is a normal workday. Canavesi reaches out to manufacturers to answer their questions and explore new opportunities in the market. As he heads out into the parking lot, a sea of stationary vehicles surrounds him. Most of them differ from each other and are all awaiting the return of their owners for the drive home.

Canavesi, his steps brisk and purposeful, shares an insight that hints at a transformation of an everyday occurrence. "You know, all this could look a lot different,” he says, sweeping his hand across the landscape of parked cars.

His assertion is in reference to new technologies in safe autonomous driving. "So much mobility is so time-consuming right now,” Canavesi notes, moments before slipping behind the wheel of his own vehicle. "Improving sensors and enabling autonomous driving would give people time to do other things while commuting. Imagine what people could accomplish if they could focus on other things in the car instead of driving.” With that he closes the car door and drives off to end another day trying to redesign what we see as normal.

In focus: autonomous driving

  • Self-driving cars, a key component of future mobility championed by ZEISS, rely on machine learning and artificial intelligence to enhance safe autonomous driving. The expertise of ZEISS is instrumental in manufacturing the chips that enable autonomous vehicles to interpret complex data for smarter navigation and decision-making, leading to safer mobility. By reducing human errors, the principal cause of road incidents, advancements in driverless car technology promise to make our journeys more reliable and efficient.

  • The latest innovations in autonomous driving are focused on enhancing safety, reliability and integration with smart city infrastructure. The ZEISS spin-off Scantinel Photonics is, for example, at the forefront of developing second generation “light detection and ranging sensors” (LiDAR). This technology for optical distance and speed measurement is vital for empowering autonomous smart systems to see as well as, if not better than, humans. The company focuses on "frequency-modulated continuous wave technology" (FMCW) to create significantly advanced new LiDAR sensors. With these sensors, distances and speeds of the detected objects can be measured with greater accuracy, significantly improving the performance and safety of autonomous applications.

    The manufacture of the new optical sensors is based on photonic integration, in which special optical components for processing light pulses, similar to electronic circuits, are applied to a silicon chip. These innovations collectively represent a significant leap toward improving self-driving cars. Scantinel’s LiDAR on a chip, with Photonic Integrated Circuit (PIC) design, is set to revolutionize autonomous mobility by enabling versatile integration capabilities, much like how integrated circuits transformed electronics by condensing complex functions into compact, efficient components.

  • Fully automated driving technology aims to relieve human drivers of the burden of driving tasks, utilizing advanced driver assistance systems and higher levels of vehicle automation. Driverless cars, equipped with automated driving systems that harness machine learning and artificial intelligence (AI), promise to mitigate common road traffic issues such as traffic jams and accidents caused by human error.

    Safety drivers are currently used in testing to ensure the technology can handle a variety of traffic situations safely, while the automotive industry works toward achieving a reliable automated system. As these technologies evolve, they have the potential to significantly reshape transportation and increase the safety of all road users. Despite incremental advancements, achieving the widespread adoption of fully automated, driver-free vehicles that can handle complex driving environments reliably remains an ongoing challenge.