1. What are the key trends in the ADAS and AD space from an industry perspective? How is LeddarTech addressing these trends?

The automotive industry is witnessing increased automation, connectivity, and electrification (ACE) trends. LeddarTech addresses these by providing comprehensive perception software solutions essential for ADAS and AD development.

LeddarTech focuses on enhancing the efficiency and safety of these technologies by offering scalable and flexible software like LeddarVision. This allows automakers to adapt to evolving industry trends seamlessly.

For decades, the automotive industry assumed car crashes were inevitable. Now, the industry is working to avoid them through legislation (like GSR 2022, for example). Today, almost 95 percent of accidents are caused by human failure, three quarters of which are due to problems related to staying in a lane. Regulators are now mandating that car manufacturers overcome human shortcomings with automated driver assistance. Current vision-based systems, which use object-level fusion to analyze camera data, are still insufficient. An independent analysis in 2020 determined that these safety systems malfunction once every 8 miles and often fail to meet basic requirements, such as pedestrian identification. Vision-based architecture will struggle to meet the regulations of 2025 and beyond at reasonable costs.

Cars are becoming more complex and outfitted with multiple sensors and processors to address this challenge. But how they process the data from these sensors matters. Object-level fusion systems currently used by most cars interpret each sensor independently and often have trouble resolving inconsistencies.

LeddarTech’s raw-data fusion software combines the output of all sensors to create real-time, 3D models of a car’s environment. Our more accurate understanding of a car’s surroundings improves automated decisions. Lane detection-assisted cruise control, automatic emergency braking or acceleration and pedestrian avoidance are all improved by LeddarTech’s pioneering software. Our software stack offers 2x the performance at almost 50% lower sensor cost.

  1. With the industry favoring vision-based solutions, how do you think the market will evolve, especially for radar and LiDAR?

The automotive industry’s preference for vision-based solutions does not necessarily mean a decline in the importance of radar and LiDAR. Instead, these sensors are evolving rapidly to complement vision systems. Radar and LiDAR are still vital components, albeit LiDAR is more valuable at higher levels of autonomy. Radar and LiDAR offer unique capabilities, and their integration with vision sensors drives the development of more capable and safer autonomous vehicles. The future likely holds a landscape where sensor fusion, incorporating radar, LiDAR and vision, plays a pivotal role in ensuring the safety and reliability of autonomous vehicles.

Today, basic ADAS warning systems primarily use vision-based systems. However, regulations and consumer demands for greater safety and convenience contribute to the acceleration of integrating multi-sensor ADAS systems that include cameras, radars, LiDARs and ultrasonic sensors.

Some high-performance premium ADAS can depend on more than 20 sensors mounted at various locations of the vehicle. Each sensor provides different types of data, such as cameras provide images, radars and LiDARs provide point clouds, etc. The LeddarVision software stack fuses the raw data from each of these sensors individually and collectively from multiple sensors (radar + camera) to make precise driving decisions through the most robust and reliable environmental and perception software. This perception is critical, enabling vehicles to make the right decisions, i.e., stop, accelerate, turn, reverse, etc.

At higher levels of autonomy (L2+ and above), where autonomous features replace driver tasks, multiple sensors are required to understand the environment correctly. Multi-sensory systems are necessary given that every sensor is different and has limitations, i.e., a camera will work very well for lane detection or object classification. In contrast, a radar may provide good data for long-range detection or in different light conditions. The LeddarVision software stack enables the highest performance through low-level (raw data) fusion and perception systems by improving the performance of each sensor and through a combined unified environmental model.

  1. There is a lot of talk about software-defined vehicles (SDVs); what could be its impact on ADAS and AD?

 LeddarVision’s software-defined approach aligns with this trend, enabling automakers to update and customize perception software without hardware modifications. This flexibility is crucial for the evolution of SDVs as it allows for continuous improvements and adaptations.

As vehicles become more software-driven, there is increasing pressure on OEMs and their suppliers to write, deploy and integrate code more quickly and efficiently. The shift to SDVs profoundly affects every function and facet of the automotive industry. Not only is software playing a larger role in core vehicle functions, but it is also enabling new features such as advanced driver assistance systems and autonomous driving. OEMs need to incorporate new features closer to the start of production, and they also need the ability to quickly and safely push out software upgrades after vehicles have been produced.

Traditionally, OEM developers system and integrators have written software for each hardware component and then integrate it with code for other vehicle parts. Testing the integrated software has come late in the process, limiting the time for making additional changes. Developing each component and vehicle platform has been a one-off process that starts over again for the next platform. This is not scalable. As a result, OEMs are shifting from this vertically integrated approach to iterative methods that incorporate agile software development, collaboration with software vendors (like LeddarTech), integration and code reuse. For example, the LeddarVision platform, which delivers best-in-class low-level fusion and perception solutions as a horizontal layer of middleware, enabling OEMs and Tier-1s to develop applications and functions integrated via a robust application programming interface (API) and shared among the development teams and offering unique industry scalability and reuse! The LeddarVision software stack helps OEMs and Tier-1s to implement new ADAS architectures and features scaling from entry model to mid to premium, leveraging the same unified environmental model software, reducing cost and complexity.

  1. What could be some major roadblocks for software developers in the automotive industry over the next decade?

 Some challenges for software developers in the automotive industry include stringent safety and cybersecurity requirements.

Meeting safety standards like ISO 26262 and ensuring the security of connected vehicles are critical. LeddarTech addresses these challenges by providing robust, NCAP and GSR safety-compliant software solutions.

One of the most significant changes to automotive system developers is the adoption of agile software development cycles, which are designed to be more flexible and adaptable than the traditional sequential waterfall development processes common in automotive manufacturing. Shorter development cycles, frequent releases and a focus on collaboration and continuous improvement characterize agile development processes. This approach is well suited to software development, as it allows teams to respond quickly to changing requirements and market conditions and to incorporate feedback from users and stakeholders.

However, moving towards agile software development cycles, such as the V model, can conflict with traditional hardware development methodologies, creating a culture clash between the software and hardware teams. For example, teams that typically work in a structured and sequential manner may struggle to adapt to the more flexible and collaborative approach of agile development. Conversely, teams working in an agile development environment may struggle to fit into traditional hardware development’s more structured and hierarchical setting.

  1. LeddarVision talks about scalability and modularity; can you explain this in more detail?

While scalability typically refers to the ability to adapt to different levels of automation (from ADAS to highly automated driving), LeddarVision’s modularity extends this concept. It allows automakers to scale computational power and sensor configurations effortlessly. Developers can integrate new sensors into the existing modular framework as sensor technologies evolve without extensive rework.

LeddarVision’s modularity is a key feature that sets it apart in the ADAS and AD space. Modularity refers to the software’s ability to efficiently handle various use cases, features and sensor sets while maintaining a unified architecture. This modularity advantage benefits automakers and developers by enabling customization, reducing development efforts and future-proofing. LeddarVision’s modularity future-proofs perception systems. When new features or safety standards are introduced, developers can integrate them seamlessly into the existing modular framework. This adaptability ensures that vehicles equipped with LeddarVision remain competitive and compliant over time. LeddarVision’s modular architecture also facilitates interoperability with various hardware platforms, allowing them to choose from a range of sensors and processors while maintaining a cohesive perception system. It simplifies integration and allows for hardware upgrades without requiring extensive changes to the software.

  1. Partnerships are vital to the development of solutions in the ADAS and AD space. Can you take us through some of the factors that influence the selection of partners for an OEM and a Tier 1 supplier?

Partnerships are crucial for success in ADAS and AD development. Factors influencing partner selection include collaboration, innovation, the ability to be flexible and integrate OEM and Tier-1 innovations and inputs in development projects, automotive technical expertise and experience with functional safety, open innovation, financial capacity, and scale to support long-term projects, and the ability to complement each other’s strengths.

  1. Continuing the topic of partnerships, can you provide a few examples of LeddarTech partnerships and the impact that these partnerships are expected to make in the automotive ecosystem?

LeddarTech collaborates with partners who share its vision of advancing automotive safety and efficiency. Such partnerships leverage LeddarVision’s sensor fusion and perception capabilities to create comprehensive ADAS solutions.

I can share just two of many examples of our collaboration. One is our commercial agreement with Ficosa, which involves the successful integration of LeddarTech’s advanced LeddarVision software with Ficosa’s cutting-edge surround-view camera system, along with other sensors such as radars, IMU and GPS. LeddarTech and Ficosa continue to strengthen their strategic collaboration towards developing a pioneering Intelligent Parking Assistance system, allowing even more detailed and precise perception of the surroundings.

Another example is our collaboration with Texas Instruments on our LeddarVision entry-level ADAS L2/L2+ highway assist and 5-star NCAP 2025/GSR 2022 low-level fusion and perception software stack, which we optimized with TI’s TDA4VM-Q1 (8 TOPS) processor which achieves one of the lowest system costs for L2/L2+ entry-level ADAS without sacrificing system performance.

These examples illustrate how real collaboration partners can work together to address and help solve automotive OEMs and Tier 1s’ pain points, including cost, performance, safety and scalability.

  1. In general, what is your outlook for the upcoming year, especially for LeddarTech?

I am very optimistic. By the end of 2023, we will have released three new software products and have gained considerable traction with OEMs and Tier 1s. In addition, we announced a business combination (SPAC) with Prospector Capital in June and plan to be publicly listed by the end of the year. This is a very important moment for LeddarTech as it will enable the company to scale faster to serve more customers.

For 2024 and beyond, we will continue to release more software features, expand our processor and sensor support list, engage with more OEMs and Tier 1s and expand our partnerships. We plan to continue to grow the organization in all our current locations.

About Thirumalai Narasimhan

Thirumalai is an Automotive engineer with ~10 years of experience, working on Body Systems Design, with OEMs like Renault-Nissan and Mercedes AMG, having worked extensively on bringing concepts to the production line in various passenger vehicles. In his current role, Thirumalai works with the Chassis Safety and Autonomous Driving team, leveraging his industry experience to bring insights into Research reports.

Thirumalai Narasimhan

Thirumalai is an Automotive engineer with ~10 years of experience, working on Body Systems Design, with OEMs like Renault-Nissan and Mercedes AMG, having worked extensively on bringing concepts to the production line in various passenger vehicles. In his current role, Thirumalai works with the Chassis Safety and Autonomous Driving team, leveraging his industry experience to bring insights into Research reports.

Your Transformational Growth Journey Starts Here

Share This