Had time to do more DD this AM - this is what I was looking for, sensor fusion...
1. Intelligent Modular Architecture
This group of patents relate to AEye’s intelligent modular bistatic architecture, system design, and its solid-state performing MEMs-based agile LiDAR. While many other LIDARs process in a linear or serial manner, AEye parallel processes using a bistatic architecture. This approach allows AEye to transmit and receive light out of separate paths, enabling the delivery of its iDAR™ perception system and adaptive capabilities that power its industry-leading range, resolution, refresh rates, agility and intelligence. Equally important, AEye’s design allows for flexibility in hardware packaging and vehicle placement options, unlike some LiDAR sensors that are limited to roof placement. These patents include protection of the overall architecture, of feedback techniques to keep the scanning where desired during vehicle vibration and temperature variation, as well as the ability to control detection pulse width and receiver bandwidth.
2. Open System and Extended Data Capabilities
A second group of patents focus on AEye’s software definable AI technology and iDAR’s ability to integrate other existing sensors, such as radar, cameras, and IMUs. These patents address AEye’s unique co-boresighted design, wherein an HD camera and LiDAR receiver share the same optical axis, creating true color point clouds which enhance classification capabilities and eliminate the need for post-processing parallax correction. By co-boresighting the camera with the LiDAR receiver, AEye moves intelligence into the sensor. This enables the sensor to detect and immediately request additional data about objects and anomalies, then issue “fast path” priority message alerts to inform the motion planning system. By pushing detected threats to the top of the data stack under consideration by the motion planning system, AEye reduces central processing burdens, and speeds time to reaction.
3. Enhanced Data Quality
The third group encompasses innovative features such as scan agility, which enables the sensor to “acquire” pre-classification attributes useful in accelerating perception systems. For example, AEye is able to pinpoint intra frame velocity by capturing both radial and lateral velocity, where most threats occur while driving. These patents build a software-defined layer upon the bistatic platform, and include the ability to optimize shot lists as a function of scenes and environment, novel tracking schemes, and tracking the horizon during car motion. The latter is an important aspect of agility and intelligence, as it reduces the required vertical scan field of view, and therefore reduces frame to frame time. iDAR’s ability to enable software-defined frames and dynamic scan patterns and the platform’s ability to deploy adaptive energy control on a pulse-by-pulse basis enables both the dynamic adjustment of scan patterns and the ability to adapt the laser energy for each pulse for complete interference mitigation.
4. Optical Data Networking
The fourth group enables AEye to expand the capabilities of the AEye sensor to enable the laser to communicate information. This groundbreaking feature is uniquely enabled by AEye’s adaptive LiDAR, in which the laser can be directed with precision. Pulsed message packets enable data to be transferred via line-of-sight. By using optical communications via the LiDAR system, AEye sensors can leverage a readily available, secure and reliable communications channel, which does not compete with the congested bandwidth of WiFi, cellular, and/or satellite communications, to transmit, receive, and/or transceive data.
“We set out to build a flexible architecture that could be expanded and adapted over time, adding new innovations. Our IP strategy, we believe, places us years ahead of others building single use hardware,” said Dr. Allan Steinhardt, chief scientist at AEye. “We will continue to push forward aggressively with developing and protecting competitive intellectual property to advance AEye’s technological leadership. We are creating a thorough and rigorously defensible patent portfolio covering the architecture and software that underlie our solution.”
AEye’s iDAR™ system has been independently verified to have significant range, resolution, and speed performance advantages, and is software-configurable to serve multiple markets, including automotive, industrial, and mobility, with the same platform and supply chain. The company has recently announced technology and/or manufacturing partnerships with NVIDIA, Sanmina, Continental, Benchmark, and TuSimple, as well as its expansion into the Japanese and Korean markets as it prepares for volume production.
To get the latest AEye news and technology, sign up for the company newsletter at aeye.ai/updates-sign-up/ or to see these many of these latest innovations in action and to experience iDAR’s record-breaking performance in real-time, visit aeye.ai/demo-the-4sight-m.
About AEye
AEye is the premier provider of intelligent, next generation, adaptive LiDAR for vehicle autonomy, advanced driver-assistance systems (ADAS), and robotic vision applications. AEye’s iDAR™ (Intelligent Detection and Ranging) system leverages biomimicry and principles from automated targeting applications used by the military to scan the environment, intelligently focusing on what matters most, enabling faster, more accurate, and more reliable perception. iDAR is the only software configurable LiDAR with integrated deterministic artificial intelligence, delivering industry-leading performance in range, resolution, and speed. The company was founded in 2013 and is based in the San Francisco Bay Area.
