Solid-state LiDAR technology being fine tuned in the field

Solid-state 3D system-on-a-chip LiDAR (light detection and ranging) have moved out of the laboratory and into the development departments of OEMs and Tier 1 companies.

Evaluation kits are being provided by Quebec-based LeddarTech. “We’re proud to be the first company in the industry to deliver 3D solid-state LiDAR SoCs (system-on-a-chip),” said Michael Poulin, LeddarTech Vice-President, Product Management in a press statement. Recent reports from industry analysts on LiDAR market opportunity confirm the strategic position of Solid-state LiDAR (SSL) technology in terms of enabling mass-market autonomous driving deployments.

Infinium Global Research predicts that there will be a compound annual growth of 35% from 2017 to 2023. Another research company, Technavio, has found that LiDAR is the sensor of choice for automotive driving applications, with industry experts confirming its necessity for self-driving cars. Transparency Market Research came to the same conclusion. It has released findings that predict the demand for solid-state technology will support a multibillion-dollar automotive LiDAR sensor market, taking market share from mechanical scanners. And, in its recently released market report overview, technology and market analyst firm Yole Développement expects the global automotive LiDAR market to reach US$5 billion by 2023 and US$28 billion by 2032.

Automotive Industries (AI) asked Poulin how innovative the company’s LiDAR solution is compared to similar technology.

Poulin: SSL technology has been identified as a key enabler for mass-market, self-driving vehicles. While mechanical scanning LiDARs deliver interesting performances and are being used in many autonomous driving R&D vehicles and early prototypes, their complex build makes them overly expensive and lacking in robustness for production automotive deployment.

Our LiDAR development platform brings solid-state LiDAR technology from discrete designs to systems-on-chip (SoCs). Our patented SSL technology generates unmatched sensitivity and makes it possible to produce LIDAR sensors that achieve the automotive industry’s targeted performance levels using more affordable, readily available components—as well as simple, robust sensor designs with no moving parts.

One implementation is flash LiDAR, which means that the LiDAR emits broad light pulses that illuminate the whole area of the field-of-view, similar to a camera flash. Other LiDARs use a point-scan of collimated laser beams and mechanical scanners that only capture a limited number of points on each object, with gaps in between points.

The powerful, patented Leddar digital sig­nal processing is also a key differentiator. Many approaches to designing LiDAR rely predominantly on hardware-based meth­ods to implement detection and ranging, providing raw data or simple readings. In comparison, the Leddar technology relies primarily on software and signal processing embedded into the core of each sensor. This al­lows for rapid delivery of accurate measurements, extends the sensor’s effective range, and provides superior lateral discrimination and capability to track multiple objects simultaneously, even in challenging weather condi­tions such as rain, snow, or fog.

AI: How will the LeddarCore LCA2 advance autonomous driving?

Poulin: The LCA2 is our first SoC offering optimized specifically for the needs and requirements of the automotive market. It enables the design and mass production of solid-state LiDARs at very competitive prices – typically a few hundred dollars in volume – that meet OEM short-term requirements for mid-range applications, such as corner LiDARs or traffic jam chauffeur. The upcoming generation called the LCA3, due for mass availability in 2020, will provide even more range and resolution, which are required for fully autonomous systems. But perhaps the most significant advancement brought forth by the LeddarCore offering lies in its underlying business model, which provides significant added value to the automotive industry. Unlike most LiDAR vendors, who manufacture and sell one-size-fits-all LiDAR sensors as a finished product, LeddarTech provides a LiDAR development platform so that Tier 1 vendors can develop and produce customized LiDAR solutions.

LeddarTech has also developed an ecosystem of world-class suppliers who provide components such as emitters, receivers, micromirrors, microprocessors and software development tools that are prequalified for integration with the LeddarCore SoC, improving LiDAR design efficiency and agility for our clients.

AI: What is the purpose of your LCA2 evaluation kit?

Poulin: The LCA2 evaluation kits are complete, fully functional LiDAR sensors built with the LCA2 SoC. These evaluation kits accelerate and facilitate the evaluation and testing of the LCA2 SoC by our automotive clients and partners, and they support the immediate development of applications and software that will be using future designs of LiDARs based on the LCA2.

AI: What are some of the challenges facing automotive OEMs with regards to using LiDAR?

Poulin: Let’s look at three of the fundamental challenges to implementing automotive-grade LiDARs in production vehicles to enable autonomous driving.

First, the cost/performance ratio. OEMs are looking for the most cost-effective LiDAR technology which best meets the performance requirements of the intended application. The Leddar technology enables the development of LiDARs that can use more affordable and readily available components to achieve the same level of performance as more expensive systems.

Second, the industrialization process. Many emerging LiDAR vendors propose sophisticated technologies that have never been manufactured on a large scale before. Yet, going from the labs to mass commercial production is a complex process that involves significant risks in terms of timeframe, volume, quality, supply chain, etc. We believe the accelerated production ramp-up and mass availability of automotive-grade LiDARs will be key for OEMs. This is why we built a business model where we provide the core LiDAR expertise and team up with multiple Tier-1 manufacturers.

The sensor must be automotive-grade and must be able to perform reliably for years in the harsh conditions faced by automotive components. Our development platform is based on demonstrated solid-state LiDAR designs and components, and benefits from the integration of many discrete functionalities into single SoCs. LeddarCore SoCs and related software library are being developed in compliance with ISO-26262 ASIL-B functional safety norms.

AI: What will LeddarTech be showcasing at CES Las Vegas 2019?

Poulin: Our presence at CES will focus on the LiDAR development platform, including our LeddarCore SoCs. Visitors will be able to view demonstrations of LiDAR designs based on the LCA2 SoCs, as well as a new LCA3 LiDAR solution for long-range applications. Our pavilion will also feature client solutions and a dozen members of the Leddar Ecosystem, who will demonstrate their solutions. Finally, the pavilion will have a seminar area where guests will be able to attend informative sessions relating to LiDAR development and implementations.

AI: What industry partnerships is LeddarTech pursuing for intelligent transport systems?

Poulin: The ITS market is entering a transformational period where sensors, intelligence and automation are seen as a way to optimize existing and future infrastructure, improving transportation efficiency for both goods and people, and positively impacting the environment and our quality of life. These applications typically require the integration of sensors into applications that need to be customized to specific development scenarios.

Hence, LeddarTech usually works with companies that specialize in the integration of sensors into various traffic management applications, or with large engineering firms developing complete ITS solutions that require the efficient detection of vehicles, cyclists, or pedestrians. ITS (Popular ITS applications) for LiDARs include e-tolling, vehicle profiling, and stop bar detection for traffic light automation.

AI: How do you see artificial intelligence and mobility coming together?

Poulin: Automation, in both vehicular and traffic management applications, will require significant AI to reach their full potential. Both vehicles and ITS will need to share and process information in real time to achieve optimal efficiency and provide maximum added value for users and commuters. Moving to this intelligent mobility model will also enable the emergence of many new applications and business models. This will not happen overnight, and we are still only at the onset of mobility automation, but we can expect it to have a profound, transformational impact on the automotive industry and the way we use transportation in the future.