LiDAR: Intel RealSense L515 LiDAR Camera

--

TechInsights publishes a teardown of Intel RealSense L515 LiDAR:

“The Intel RealSense L515 is rated for indoor use and can provide depth measurements between 0.25 m and 9 m with a field of view of 70° x 55°. Its LiDAR unit consists of an IR (860 nm) laser reflecting off a scanning MEMS mirror, coupled with an IR photodiode. It is also fitted with an RGB camera which provides a 1080p resolution at 30 fps.”

The world of LIDAR sensing is evolving. Rotating turret LiDARs are commonplace for applications such as autonomous driving (see our Automotive LIDAR teardown subscription), but they are being displaced by a new generation of solid-state laser scanning LiDAR technology.

Of particular interest to us is the recently released Intel RealSense L515. It is targeted at warehousing & logistics, industrial robotics, and indoor 3D scanning applications. The combination of LiDAR and Camera into a single module and thereby internalization of sensor fusion made it an interesting device to analyze. We were eager to get our hands on one of these units and proceed with a full teardown (see our complete report here).

Specifications

The Intel RealSense L515 is rated for indoor use and can provide depth measurements between 0.25 m and 9 m with a field of view of 70° x 55°. Its LiDAR unit consists of an IR (860 nm) laser reflecting off a scanning MEMS mirror, coupled with an IR photodiode. It is also fitted with an RGB camera which provides a 1080p resolution at 30 fps.

The RealSense L515 is housed in a compact and lightweight “hockey puck” enclosure (61 mm x 26 mm / 100 grams), which makes it ideal for robotics applications, as it can be more easily incorporated into a product.

RealSense L515 Teardown

Once open, we were able to identify the following design wins:

FunctionManufacturerPart Number

The Intel RealSense L515 LiDAR camera is based on two major ICs from Intel, the RealSense vision processor, and the MEMS controller. The power management chip was supplied by Analog Devices.

The 2 MP camera was delivered by OmniVision with the OV2740 image sensor, which also operates as the color sensor. This device also contains an infrared laser projector EEL operating at 860 nm. These optics utilize an advanced stereo depth algorithm for accurate depth perception at a long range (9 m).

This device does not contain an internal power source; it is powered by a USB type-C port.

Of particular interest is the MEMS mirror used in this LiDAR camera. The following section explores this component further.

MEMS Design Win: STM PM56A

The optical board hosts 2 major components of the LiDAR: the edge emitting laser, and the MEMS mirror die mounted between 4 permanent magnets.

TechInsights was able to remove an intact STMicroelectronics PM56A die from the laser scanning module. The die is measured at 3.53 mm x 2.8 mm.

Compared to its predecessor the PM54A, analyzed by TechInsights, the mirror found in the Intel RealSense L515 does not have a static frame, rather all static parts (such as PADs, routing metals, anchors) are inside the maximum in-plane extension defined by the Lorentz coil. This solution allows to have a more compact design, despite the high Field of View reported by Intel of 70° x 55°.

What is the cost of the hardware in the RealSense L515?

Our costing specialists assembled the following cost breakdown for the L515.

The main cost contributor is quite as expected integrated circuits, at $16.45 on a total of just over $35. The retail price of the L515 is $349.

Conclusion

MEMS-based laser scanning LiDAR cameras can provide an effective way of reducing electronics costs while improving the overall reliability of these devices. The inclusion of this technology in the RealSense L515 allows Intel to pursue logistics and robotics markets, where more expensive and fragile technologies may not be as widely adopted. TechInsights has additional analysis available on this technology, including cross-sectional analysis of the MEMS die, which can be found here.

Originally published at https://www.techinsights.com.

--

--

For Data sensing and AI Infrastructure
For Data sensing and AI Infrastructure

Written by For Data sensing and AI Infrastructure

Marc. Y. , Product and R&D Director. Focus on Data Sensing and AI Infrastructure. More info at http://4da.tech.