Earlier in January this year, Quebec-based LeddarTech, a pioneer in environmental sensing technologies for advanced driving assistance systems (ADAS) and autonomous driving (AD), entered into a long-term agreement with Germany-based OSRAM, a global leader in automotive lighting systems. LeddarTech will provide their industry-leading LiDAR hardware and software components into OSRAM’s PERCEPT™ LiDAR platform. The PERCEPT LiDAR platform is the first flexible solid-state LiDAR platform engineered with a strict focus on industrialization and automotive qualification. In a clear Tier 2 role, OSRAM wants to offer a mid- to long-range platform to OEMs, Tier 1s, and system integrators that can be adapted for their specific application.
To learn more about what the partners hope to achieve through their collaboration, we spoke with Charles Boulanger, CEO of LeddarTech.
For Osram, this partnership reflects a strategic move to carry forward its dominance in the photonics systems market into the LiDAR market. The tie-up with LeddarTech allows it to more effectively target the LiDAR space, wherein LeddarTech’s cutting-edge LiDAR acquisition and procession integrated circuit along with related measuring software technology will be leveraged into Osram’s Percept LiDAR platform.
With its competencies in perception technology, sensor fusion and LiDAR solutions, the motivation for LeddarTech to partner with Osram is to become the most deployed sensing platform in automotive and mobility domains. The overarching objective for both partners is to offer a high performance, cost-effective automotive-grade LiDAR -perception platform to Tier Is and OEMs and help them accelerate the mass deployment of AD systems.
Although it is among the most reputed solid-state LiDAR developers in the world today, LeddarTech realized early on that due to challenges related to mass production, access to OEM clients and their platform specific requirements and, most importantly, achieving cost, collaboration rather than competition would be key to future success. Trends also indicated that OEM customers were more likely to repose their trust in established Tier I companies, rather than start-ups owing to risks related to warranty and liability in case of system malfunction.
The top five companies are expected to control anywhere between 50-80% when three key technologies for AD – camera, short range and long-range radar – go mainstream and into volume production. So, the question is why LiDAR runs very differently from these categories?
In the early stages of closed innovation, the norm has been to try and develop high performance LiDAR sensors to address most perception challenges. As sensing technology and the ADAS market mature, the automotive industry has recognized that in order to achieve higher autonomy, sensors must complement each other’s performance capabilities, provide complete understanding of vehicle surroundings by operating, either in silos or fused together, and achieve perception redundancy. In this context, the issue becomes not one of sensors but of systems. “This requires the whole system to be optimized; it’s not necessarily the case that the best radar or LiDAR represents the best solution, it’s more about what the optimal solution will be,” noted Boulanger. “This means a lesser performance radar / LiDAR has to be continually tweaked to get the best system.”
Accordingly, LeddarTech’s decision to partner is rooted in the conviction that there is no one-size-fits-all, that continuous organic modifications will be required to provide actionable data, and that an innovative business model will be needed that can disrupt the entire value chain and enable industrialization to follow. Its focus is to deliver actionable data and work with Osram to address the industry demand for a cost-effective, high performance solid-state LiDAR with auto grade components and functional safety.
“Our business model is not focused on a complete black box approach but on providing a toolbox of fundamental technologies that are required by almost everybody to enhance car performance sensing.
Since then we have aggregated more technology to the portfolio in our toolbox. We have added solid-state beam steering on top of signal processing,” added Boulanger.
How it Will Work
Most OEMs are leveraging the expertise of Tier Is and working with them to finalize the sensor suite for Level 3 and above autonomy. In turn, Tier I’s are working with Tier II suppliers and technology developers to identify and build new sensor capabilities to complement their in-house sensor offering that can address a specific level of autonomous driving. OEMs are depending on Tier Is to support them in achieving a certain level of autonomy with sensors, processor platforms, and software and integrating them into the vehicle.
“The Osram solution, even for Tier Is, will be a good way to create a product which they can customize, add value, integrate into a more complex system and subsequently enhance with software so it accelerates the development process,” stated Boulanger.
Short-term vs Long-term Strategies
While there’s no doubt in the industry that LiDAR will be a necessity, it’s also very clear that Tier Is see the biggest opportunity over the next few years being Level 2 / Level 2+ and, indeed, by 2030 that’s where the volumes will be. Level 2 / Level 2+ will be mainly radar and camera, with increased market deployment of LiDARs coming in only post-2025 with Level 3 and above autonomy. From Frost and Sullivan’s perspective, a few premium vehicle manufacturers are likely to choose LiDAR based sensor suite solution for their Level 2+ offering which will enable scalability to Level 3 and Level 3+ using OTA.
”Our competitors in the LiDAR market are pushing a product, not a solution,” explained Boulanger. “In contrast, our wins are all platform wins; our customers can use our platform to develop a whole host of products with it. So, in a sense, the short-term game is for our competitors, the long-term game is for us. Simultaneously, since we’re not only LiDAR and because our sensor fusion supports radar and camera-based systems, we are able to tap into the camera/radar ADAS market, i.e., the Level 2+ market which represents a more short-term opportunity.”
We were also keen to learn about how LeddarTech-Osram planned to engage with the rising trend of features on demand? For instance, how they planned to approach the idea of over-the-air (OTA) updates to ensure that the vehicle could achieve Level 3 and Level 3+, for example, offering redundancies and enabling high-speed and nighttime driving,
Development cycles in hardware tend to be lengthy and if an OEM wants product improvements on a regular basis, they need a suite that can be almost instantly improved with software OTA. Hardware cannot be improved at the same speed because of standards, certifications and regulations. In fact, just acquiring the components takes lead time whereas in software the improvements are immediate.
“That’s part of our business model: our solutions are software-centric,” elaborated Boulanger. “From the outset, we adopted a software-centric design to create the flexibility but also extensibility of LiDARs and perception systems, including through OTAv upgrades.“
A Palette of Different Applications
We were intrigued about whether the LiDAR market would target application segments like robotaxis, curbside delivery, last mile delivery, and food delivery until Level 3 becomes more mass market and OEMs capture a bigger chunk of LiDAR. Boulanger asserted that there were indeed two target segments: first, passenger cars and, second, mobility that encompassed shuttles, delivery, commercial trucking, off road, and robotaxis. However, he was also convinced that these would represent only limited volumes in the short-term. The volumes that enable optimization of cost, size, power required are just not there, with sales only in the thousands.
Tier I’s will be selling to this market as well but not in the short-term. Today, the interest is in demonstrating that the system can work and that a car can be autonomous. Many companies are using brute force, placing multiple LiDARs, cameras and radars on the car. This generates challenges related to maintenance since any possible misalignment could immediately result in the car being sent for lengthy, expensive recalibration. While this might still be feasible when it is only a question of a few cars, the real challenge emerges when subsequent scaling is required. Therefore, resolving this problem is critical if scale is to be achieved. The market for LiDAR modules will be tremendous with both markets ultimately merging over the long-term.
Full perception software stacks:
While the automotive industry has accepted LiDAR as a critical sensor required to achieve higher autonomy, they have comprehended that achieving complete perception is beyond a single sensor or even multiple sensors working in silos. To address this challenge, LeddarTech has extended it’s offering to LeddarVision, an open software platform that combines AI and computer vision technologies to scale up the performance sensors and hardware. LeddarVision is a hardware and software agnostic, open sensor fusion and perception platform that provides complete environment model by fusion of raw data from sensors such as radars, LiDARs, cameras, ultrasound and GNSS sensors, scalable form L2 to L5 autonomy.
Frost & Sullivan Perspective
In a generic value chain, where LiDAR companies partner with Tier I suppliers to develop and industrialize application-specific LiDAR systems, LeddarTech has collaborated with OSRAM, another Tier II supplier (or what we, in Frost and Sullivan, call a Tier 1.5 supplier who supplies to both Tier I and Tier IIs) to offer a flexible solid-state LiDAR perception platform.
This collaboration and shift in value chain will allow both companies to combine their respective technologies and expertise in LiDAR, perception sensing, photonics and industrialization to achieve low cost, customizable LiDAR solutions with reduced risk of warranty and liability, while enhancing the value proposition to OEMs/ Tier Is/ system integrators to leverage on the larger price margins, making this collaboration more sought after.
Whilst most LiDAR companies are developing LiDAR products based on a specific combination of optical and hardware components, the LeddarEngine™ platform allows the use of multiple different technologies and components such as Time-of-Flight (ToF), MEMS mirror, solid-state beam steering, APD, SiPM and many others, making it truly flexible and versatile at a time when the industry is working to realize a cost-effective, high-performance LiDAR.
Focusing on Level 2+ and above applications, the jointly developed flexible platform will allow Tier I customers to customize their LiDAR product to cater to the needs of OEM customers.
While most Tier I suppliers are sourcing LiDAR from start-up companies, this collaboration led by OSRAM will offer trust and assurance to OEM and Tier I customers on the automotive grade, quality, and performance capabilities of their LiDAR solution.
Solutions like LeddarEngine and LeddarVision have supported LeddarTech’s evolution into a leader in the perception industry, addressing every customers’ need for solutions that are hardware agnostic, scalable, and adaptable to any vehicle/ platform with varied sensor configuration.