Better LIDAR in Autonomous Systems Starts with MEMS

Figure 1. Omnitron’s MEMS step-scanning mirror handles the high-vibration, wide temperature variation required for LIDAR to function in the automotive environment. (Image: Scharfsinn86/iStock)

The global market for automotive LIDAR is expected to grow from $332 million in 2022 to more than $4.5 billion by 2028. That’s solid market growth, particularly given the decades-old challenges of commercializing LIDAR that would be affordable for automotive designs. We interviewed Eric Aguilar, co-founder and CEO of Omnitron Sensors, Los Angeles, CA, to learn about a new MEMS scanning mirror that could accelerate the market adoption of LIDAR.

Tech Briefs: Where is LIDAR today? And why hasn’t it become as ubiquitous as cameras?

Eric Aguilar: Though LIDAR’s implementation lags far behind more mature vision technologies such as cameras and radar, each of these vision technologies has its pros and cons. Cameras lack depth, which is essential to understand the environment. And they can’t see at night, which slashes their efficacy. But cameras also provide color information, which LIDAR and radar don’t support. If we didn’t have cameras to differentiate a red light from a green light, we couldn’t aspire to autonomous navigation in cars.

Radar also has its limits. First and foremost, it lacks resolution, which means that it can’t differentiate a car from a person. Yet, radar is especially valuable in bad weather because it will continue to function, while LIDAR can be affected adversely by rain or snow.

Complementing both cameras and radar, LIDAR offers its own unique attributes. It provides depth information and functions seamlessly at all levels of light because it uses an active light source. It also delivers excellent resolution, so it can perceive both moving and stationary objects.

Together, cameras, radar, and LIDAR provide the core technologies for autonomous-navigation systems, whether those systems are in a car, a delivery drone, or a robot on the factory floor.

Tech Briefs: Given the necessary role of LIDAR in autonomous navigation, what’s prevented this technology from reaching its full potential?

Aguilar: Historically speaking, LIDAR’s main pain point has been its cost. That’s because it’s a complex technology that requires great precision in terms of alignment and calibration. When you’re facing precise integration of a lens, a mirror, a laser, and a detector, you have to align to sub-micron tolerances. While the materials cost of LIDAR is acceptable, this active alignment step is where you spend all your effort, and it’s become cost-prohibitive.

Figure 2. Omnitron’s MEMS mirror meets the performance and price demands of LIDAR in autonomous drones. (Image: Mih_alex/iStock)

Also, the automotive environment is notoriously tough. It’s high vibration so it shakes the LIDAR. It’s also subject to temperature fluctuations of as much as 20-30 degrees a day. This causes continuous expansion and contraction of the optical system.

I experienced this first-hand when I led engineering teams responsible for autonomous navigation. In some cases, tiny fluctuations of even a micron would cause reliability issues in the LIDAR.

Tech Briefs: Are there any workarounds to mitigate this issue?

Aguilar: The industry employs a variety of approaches. One brute-force method involves moving only the mirror because that is the smallest component in the LIDAR system. It reduces the complexity of the alignment needs. The next step is to put this mirror on a galvonometer (galvo) with a sophisticated encoder. That approach is extremely robust, so it addresses some of the reliability issues. But it’s not compact or cost-effective. It still costs $2,000-$5,000 to get that LIDAR out the door, and that’s too expensive for adoption to the vast majority of automotive, drone, and robotics markets.

In addition to the high cost, you can still run into problems with this method. At one of the companies where I worked, we found that the LIDARs using these galvos only lasted a few months before they got out of spec or simply stopped working.

Tech Briefs: Where does Omnitron Sensors’ technology fit into the mix?

Aguilar: Based on my experience with OEMs integrating LIDAR into their systems, we knew that our customers would need a big, fast, robust mirror that does step scanning. And we knew that our mirror would need to comply with frequency modulated continuous waveform (FMCW) technology.

FMCW LIDAR systems require precise, stable targeting to accurately measure distances and velocities. Step scanning allows the LIDAR to maintain focus on a specific region or target for a sufficient period, termed “dwell time,” enabling it to collect detailed data. In contrast, spinning mirrors, due to their continuous motion, are unsuitable for the precise measurements FMCW LiDAR aims to achieve.

Our topology enables us to build a large 15 mm-diameter mirror with tens of degrees of motion and the ability to do step scanning. Our advanced electrostatic, capacitive, MEMS motor generates a significantly increased force, which enables the motor to move the mirror systems with the required speed and precision needed by modern LIDAR technologies.

Electrostatic MEMS motors utilize a voltage differential to create an attractive or repulsive force between charged elements. This force causes the rotor to move or rotate, transforming electrical energy into mechanical motion. The key advantage of electrostatic motors, especially in MEMS applications, is their ability to achieve fine, precise movements with minimal power consumption.

With our MEMS topology, we’ve improved capacitance per unit area by building a much bigger MEMS motor. Capacitance per unit area is crucial in the design of MEMS devices, as it represents the efficiency with which a device can store and utilize electrical charge relative to its physical size. A higher capacitance per unit area indicates a more efficient and compact design, enabling the creation of denser, more powerful chips.

Tech Briefs: Can you provide more technical detail on your step-scanning MEMS mirror?

Figure 3. Omnitron’s 15 mm-diameter mirror supports tens of degrees of motion and the ability to do step scanning. (Image: Omnitron Sensors)

Aguilar: Our approach to achieving precise and reliable step scanning in our MEMS mirror is threefold:

High-Performance Motor: The foundation of our system is a powerful MEMS motor designed to meet the toughest demands for both the speed and power necessary to move the mirror. This ensures rapid response times and the ability to handle significant mechanical loads without compromising performance.
In-Situ Feedback Mechanism: At the core of our technology is the ability to directly measure the position of the mirror in real-time. This in-situ feedback allows for immediate adjustments that ensure that the mirror’s position is accurately maintained. This capability is crucial for fine-tuning the mirror’s orientation and for compensating for any potential drifts or misalignments during operation. In-situ feedback in MEMS devices, particularly for precision applications like LIDAR, is crucial for real-time monitoring and adjustment of the system’s performance. We use capacitive feedback, which involves measuring changes in the capacitance between components of the MEMS device as they move relative to each other. This method allows for precise control and adjustment based on the device’s current state, ensuring optimal operation.
Robust Control System: To complement the physical components, we’ve developed a sophisticated control system tailored to drive the MEMS mirror to its desired position swiftly and to maintain it there reliably. We engineered our controller to withstand and adapt to external influences, such as vibrations and temperature fluctuations, which could otherwise impact the mirror’s stability and performance. Through advanced algorithms, our control system dynamically adjusts to these conditions, ensuring that the mirror remains stable and precisely positioned, even in challenging environments.

By integrating these three key elements, we provide a step-scanning MEMS mirror solution that offers high precision, speed, and reliability, making it suitable for a wide range of applications where accuracy and rapid response are paramount.

Tech Briefs: Where are you in terms of commercializing your MEMS mirror?

Aguilar: We’re verifying our process through fabrication of our mirror with our foundry partner. We’re also engaging with automotive OEMs and Tier 1 suppliers who are excited at the prospect of a large, robust, low-cost step-scanning MEMS mirror that will meet the most rigorous demands of LIDAR in autonomous navigation.

This article was written by Ed Brown, Editor of Sensor Technology. For more information, go here .