T1: Automotive LIDAR Technology Breakthroughs – Old Wine in New Bottles?
Harvey Weinberg, Analog Devices
Abstract: The concept of LIDAR has been around for over 80 years with the first practical units appearing about 50 years ago – coinciding with the development of the earliest practical LASERs. For much of the past five decades, military, aerospace, and meteorological applications have driven the development of many different LIDAR system architectures. Today, autonomous driving is the leading driver for LIDAR system development. Indeed, hardly a week goes by without a press release of some new start-up announcing that it has invented a new way to implement LIDAR that is uniquely suited for automotive use.
The fact is that almost all of these “new and innovative breakthroughs” are variations of architectures that have been investigated for military or aerospace use over the past 50 years. In some ways, this is good news for engineers as we are in a good position to understand how each technique works and what advantages and disadvantages it brings to automotive application.
In this mini-tutorial we will start by discussing why LIDAR is essential in the context of autonomous driving; explaining why RADAR and cameras, while popularly deployed today for automatic driving assistance, are incapable of enabling autonomous driving on its own. We will then move into describing the varied architectures of LIDAR. In broad strokes:
- Flash versus scanned LIDAR.
- Coherent detection versus direct detection.
- Monostatic versus bistatic optical systems.
- Particular attention will be paid to wavelength selection and how it influences every subsequent system design decision.
- In the context of scanned LIDAR many scanning methods have been proposed. We will discuss the pros and cons of mechanical scanning, mirrors, optical phased arrays, optical waveguides and solid state electro-optics for beam steering.
Rounding out this mini-tutorial we will present several case studies of publicly announced automotive LIDAR efforts from various companies. We will examine their principles of operation and compare their claims of performance and cost to what we would expect based on the information previously presented.
He is the Division Technologist for Analog Devices automotive division. He has been working as an Engineer for over 30 years. The first decade of his career was spent designing industrial instrumentation. In 1997 he joined Analog Devices as Principal Applications Engineer for MEMS inertial sensors where he worked on automotive and consumer products ranging from airbag crash sensors to the Nintendo Wii. In 2011 he moved to the automotive division of Analog Devices as their Systems Applications Engineering Manager focusing on safety, power train and sensor applications. Over the past four years his focus has been on LIDAR systems. He holds 8 sensor patents in fields varied from ultrasonic gas velocity measurement, to continuous self-test, to LIDAR signal conditioning. Harvey has spoken at dozens of conferences over the past 20 years, recently winning best paper award at Analog Devices General Technical Conference in 2015 and best presentation at ADAS Sensors 2016.