Luminar Tech was founded in 2012 by Austin Russell. Austin’s deep interest in optics and photonics led him to understand how these technologies could be applied to the automotive sector.
Luminar has 2 proprietary products: HYDRA and IRIS, both of them being perception enabled and always learning. The company has been outpacing General Motors in patent activity over the past few years which has been an IP giant within the LIDAR sector for years.
The patent activity of the company has been on a rise from 2017 onwards (5 years after its inception). Luminar’s filings indicate protection for a wide range of technologies for the LIDAR and its properties.
We have conducted a patent portfolio analysis of the Luminar technology and tried to dig deep into the enthralling tech that the company offers. For an overview on Luminar Technologies patents, head over here: 15 Key Insights on Luminar Technologies Patents
Patent portfolio analysis: LIDAR (56 patent families)
Luminar tech is a wizard in LIDAR technology. It uses laser-based imagery for the detection of road situations for assisting self-driving cars in real-time.
More than 70% of the patents have been filed in the LIDAR domain which clearly states that Luminar tech is at the forefront of developing LIDAR systems for self-driving vehicles.
The patent filed by Luminar tech revolves around the range and architecture of the LIDAR systems powered by superior perception algorithms steering the self-driving future to new heights.
‘Perception remains a bottleneck today for autonomous mobility and we quickly worked to find the most powerful sensors to make the perception task easier. That’s where Luminar comes in – the technology is clearly above the pack in terms of range and density, which is important for solving the most challenging problems in autonomy.’
~Alexandre Haag, chief technology officer of AID
Ranging (21 patent families)
The present patent cluster aims at the transceiver (US10338199B1) and detectors that are used in the LIDAR system. More precisely towards how the sensor measures each pixel distance.
A LIDAR’s performance is greatly impacted by the signal strength. For LIDAR, increased signal strength can be achieved in two ways: by sending more laser light out (US20190107606A1 ), collecting more light by increased measurement time(US10641874B2 ).
The cluster also includes patents that involve methods for protecting the LIDAR systems for enabling efficient ADAS solutions. The patent US10254388B2 for example aims at adjusting the power of the light source used by LIDAR for detecting the target objects and obstacles in a varied atmospheric condition like fog or rain.
Another patent US10684360B2 deals with protecting the detector lens using a method called off-axis illumination. The off axis illumination is a method in which the incident light strikes at an oblique angle rather than perpendicularly to the lens.
Moreover, the patent US10241198B2 discloses a method for automatic receiver calibration according to the energy of the returning signals.
Lidar Architecture (31 patent families)
The present cluster discloses inventions related to the electronics domain that deals with signaling, semiconductors, noise reduction methods, etc. Along with this, it encompasses the inventions dealing with the improved scanning methods for the mechanical systems.
For example: The patent US10557939B2 provides a method for improving the SNR in presence of the solar background noise for improved detections and measurements of the LIDAR system.
The patent US20200025928A1 provides techniques that can be used to improve the accuracy and reliability of the LIDAR systems when operating in the presence of jamming pulses.
Austin Russel’s aim was to develop a new system with improved resolution and range. Doing this meant that the company would fabricate many components(US10663585B2) in-house, rather than relying on off-the-shelf devices.
He discovered that automotive LIDAR needed 7 seconds to safely react while traveling at highway speeds. For this reason, Russel designed the Luminar LIDAR to operate in the infrared range(US20200076152A1) rather than the normal 905 nm range.
This portion of the infrared range can be used at greater power levels without endangering the eyesight.
Also, LIDAR systems designed by Luminar tech entail the use of indium gallium arsenide (US20190310368A1) for their chips rather than the conventional silicon.
This implementation has been applied on its advanced application-specific integrated circuit (US10267918B2), enabling LASERS to be implemented on the same surface as the computational element.
There is a desire to minimize moving parts in LIDAR since moving parts are more prone to errors than static objects.
Many LIDAR architectures claim to be solid-state but are found to have several moving components.
These anomalies in the architecture generally tend to create a performance tradeoff to deliver safe and useful autonomy. The patents US20180284247A1 and US10578720B2 aim towards the efficient safety features associated with the LIDAR systems.
LASER (4 patent families)
Luminar leaves no stone unturned for developing cutting-edge technology for the LIDAR-based systems. The present cluster relates to the use of different types of LASERs in a LIDAR system for efficient detection and scanning processes.
The patents US10418776B2 and US9810775B1 discloses a solid-state LASER and a Q-switched LASER respectively. These LASERS have a property of high beam quality(i.e the illumination patterns of a LASER beam propagation).
Patent portfolio analysis- Autonomous driving (23 patent families)
This section is based solely on defining the application of the LIDAR system along with the fusion of data from other sensors powered by learning algorithms to generate point cloud data for the detection of objects and perform autonomous operations.
We have divided this cluster into subgroups for a more comprehensive understanding.
SLAM (4 patent families)
The present cluster relates to the method of efficient vehicle navigation by the generation of a model (US8760499B2) of the surrounding environment of the vehicle. This is achieved by the collaboration of the sensor data to generate a point cloud by virtue of which the autonomous vehicle is able to analyze the obstacles, lanes, blind spots, etc.
The patent US10768304B2 involves the methods of processing the point cloud generated by the vehicle sensors. Another patent US10809364B2 discloses a method of analyzing and correcting the distortions related to the point cloud for determining the relative velocity between a vehicle and an object.
Sensor Fusion (6 patent families):
The cluster aims at synergy (US10677897B2) of the data provided by the camera and LIDAR for performing operations of SLAM, object detection, etc.
Cameras do not provide accurate or reliable range information and they do not image objects that are not emitting or reflecting light. However, the patent US10627512B1 discloses a method of the combination of the 3D point cloud data of LIDAR and the camera for a better understanding of the surrounding environment.
Luminar claims that its products can view for as long as 300m far off obstacles, and collaborating with the camera data can account for suitable driving operations.
Autopilot (5 patent families):
All the information collected by the vehicle’s ADAS is used for achieving the autonomy desired for self-driving. The autopilot feature of the self-driving vehicles targets hands-free driving of the vehicle.
The present patent cluster aims at the elimination of the human drivers and relying on the processing algorithms for autonomous path planning based on a set of motion parameters (US10394243B1).
For example, the patent US10481605B1 discloses a computer-implemented method for facilitating safe stopping of the vehicle based on the different paths generated.
AI (5 patent families):
The AI cluster relates to the method of training the neural network. More specifically to the method of generating the training data (US20200074266A1) and then using this for continuous learning.
It also includes the method of labeling (US10535191B2) the different objects in the images of the environment in which vehicles operate. This in turn helps in the generation of datasets that can be used to train the machine learning models and hence promote safe autonomous driving.
Camera (3 patent families)
The present cluster relates to the invention of the systems (US20190387185A1) related to thermal imaging. This process of thermal imaging is further used in autonomous navigation systems by detecting the obstacles present on the road. The patent (US10445599B1) for example discloses a thermal camera that uses infrared radiations to generate a thermal image which is further used to classify the object for scene-based modeling.
Luminar Technology in the future
Luminar technology started with designing LIDAR systems for the automotive industry. But since its IPO it has partnered with several companies that deal not only in car manufacturing but also in the trucking industry.
Luminar started with 4 partners in 2017 and now it has over 50 partners as of 2020 which clearly indicates that Luminar is the company to power passenger vehicle highway autonomy.
Russell announced that Luminar wasn’t content with being a lidar supplier, it was developing an automated driving system (ADS) with the capabilities of level 4 autonomy. The merger with a VOLVO subsidiary Zenseact will provide Luminar with a position to challenge the existing giants like Tesla, Waymo, Zoox in the self-driving discipline.
Along with this, the merger with Daimler trucks and torc and also Ike can prove to be a gamechanger in the trucking industry where Luminar could possibly enter the logistic and freight operation challenging the likes of TuSimple.
Luminar will not only be operational on roads but also planning for autonomous flights. The company has partnered with Airbus Inc. to improve the aircraft sensing and perception, thus paving the way for automatic aerospace solutions.
So on a concluding note, we can expect Luminar to be a leader in the LIDAR making and more patents related across all verticals can be expected from the company in the coming future.