High-precision LiDAR and motor drive solutions applied to robot 3D perception

0623-ArrowTimes-Rohm-Header-Image-820x410

For a robot to achieve autonomous movement, it requires 3D perception capabilities, which involve utilizing various sensors to continuously understand the robot's position in space. Among these sensors, LiDAR (Light Detection and Ranging) is highly regarded for its ability to provide high-precision position sensing. Additionally, robots also rely on motor drive systems to execute precise movements of their own body or limbs. This article will introduce the development of LiDAR technology and the LiDAR and motor drive solutions offered by ROHM.

High-precision LiDAR technology for accurate grasp of robot 3D perception

In the near future, AI robots refer to autonomous mobile robots (AMRs) that are equipped with 3D perception capabilities, enabling them to determine their position and recognize their own movements in three-dimensional space. This enables advanced autonomous mobility and various motion control capabilities for the robots. AMRs require various motion control functionalities, including confirming their own position and real-time processing of 3D perception data (such as object classification and tracking of surrounding objects). As a result, the use of LiDAR and camera technology has become the mainstream in the industry.

With the increasing volume of real-time 3D perception data, building deep learning models has become the current trend in technological development. Deep learning models are typically trained in ideal conditions, allowing for accurate state inference. However, in the actual working environment of robots, challenges arise due to incomplete sensor information and complex motion control requirements.

The LiDAR module used in robots imposes high requirements on the performance of laser diodes (LD). There are two main requirements for the LD used: firstly, the size of the light source should be as small as possible, and secondly, the beam divergence angle should be minimized. Typically, light emitted from semiconductor laser diodes has a certain divergence angle. Therefore, a collimating lens is required to transform the light into parallel light (Figure 1).

Figure 1: Influence of light source size on spot size

However, the collimating lens cannot completely transform the light emitted from the semiconductor laser diode into parallel light. Typically, the beam emitted from the lens has a certain divergence angle (θ), which is determined by the relationship between the light source size and the lens focal length (Formula 1).

Formula 1: θ ~ d / f (d represents the light source size and f represents the lens focal length)

To achieve high resolution, it is necessary to further reduce the divergence angle (as the θ shown in Figure 1) of the beam emitted from the lens. Regarding the light source size of the LD, since the divergence angle (θ) of the beam emitted from the lens is proportional to the LD's light source size (d), reducing d by 50% will also reduce θ by 50%. As for the beam divergence angle, Formula 1 indicates that a longer focal length will result in a smaller divergence angle. Therefore, it is preferable to use LDs with smaller divergence angles.

Figure 2 illustrates the difference in focal length when the LD beam divergence angles are 20deg and 25deg. Under the same lens diameter, if an LD with a beam divergence angle of 20deg is used, a lens with approximately 25% longer focal length can be selected. This way, the divergence angle (θ) of the beam emitted from the lens can be reduced by around 20%, resulting in a 20% reduction in spot size as well.

Figure 2: Relationship between beam divergence angle and focal length (comparison between 20 degrees and 25 degrees)

High-precision LiDAR solution for autonomous AI robots at work

To address these practical challenges in real working environments, ROHM has introduced a LiDAR solution comprising laser diodes with excellent edge detection capability and the ability to output high-precision point cloud data, GaN HEMTs for high-speed driving of laser diodes, and gate driver IC for driving GaN HEMTs. With an End-to-End model capable of obtaining stable 3D perception data, autonomous AI robots can be realized.

ROHM offers a diverse range of semiconductor laser diode products for LiDAR, including the RLD90QZWA, which achieves an industry-leading ultra-small light source size of 35µm × 10µm, and the high-power RLD90QZW8 with a light source size of 270µm × 10µm, among other various products (Table 1).

Table 1: ROHM LD product lineup for LiDAR

When using the product in robotics applications, the key lies in the accurate perception of the surrounding environment within a range of approximately 10-20 meters. For robot applications that require precise short-range distance measurement, the recommended choice is the RLD90QZWA, which entered mass production in 2022. This product features an industry-leading ultra-small light source size of only 35μm × 10μm and a beam divergence angle (fast axis direction) of 20deg. Compared to the previous product, RLD90QZW5 (light source size of 70 µm × 10 µm, θꞱ=25deg), the divergence angle of the beam emitted from the lens is reduced by 60%. By selecting the RLD90QZWA, not only can a lens with a longer focal length be used, but also higher-precision point cloud data can be obtained.

Reference design for GaN HEMT laser driver to accelerate development speed

Typically, LiDAR systems employ the Time of Flight (ToF) method for distance measurement. ToF measures the time it takes for light emitted from a light source to travel to a target object and reflect back to the receiver, allowing the distance to be calculated. If the pulse width is too large, the received light pulse signals can overlap, making it difficult to distinguish between multiple objects at close distances. Therefore, to improve resolution, it is necessary to reduce the pulse width.

To build a system that can obtain higher-precision point cloud data through narrow pulse signals, ROHM has developed a reference design called "REFLD002." This design combines the key component of laser diode driving, GaN HEMT(EcoGaNTM), with a single-channel high-speed gate driver IC specifically designed for GaN HEMT driving. Relevant design information for the "REFLD002" reference design has been published on the ROHM official website to facilitate its adoption and implementation.

Figure 3: Reference design "REFLD002" for GaN HEMT laser driver

To fully leverage the high-speed driving capabilities of GaN HEMT it is essential to use high-speed gate drivers for driving them. ROHM has developed a single-channel high-speed gate driver IC, the "BD2311NVX-LB (for industrial equipment) / -C (for automotive applications)," specifically designed for driving GaN HEMTs. Samples of this product are now available. The driving method employed in this IC not only maintains a high switching speed but also reduces overshoot by approximately 10%, surpassing similar products offered in the industry.

Furthermore, while regular GaN HEMTs have a gate-source voltage rating of 6V, ROHM's EcoGaNTM technology achieves a rating of 8V. With these advantages, users no longer need to worry about the damage caused by overshoot during circuit switching, making circuit design easier and more reliable.

Figure 4: Drive waveform of laser diode "RLD90QZWA"

In addition, ROHM has also made available a free simulation tool called the "ROHM Solution Simulator" on its official website. It comes with corresponding simulation circuits, allowing users to easily analyze waveform variations under different circuit parameters. This tool proves to be very helpful for preliminary design research.

Furthermore, ROHM's official website provides application guides, simulation models (SPICE models, Ray data), and PCB library data for individual products. By utilizing the reference design, conducting circuit simulations, and leveraging the available product data, users can significantly reduce design and evaluation time, thereby shortening the product's time-to-market cycle.

Compact and efficient motor driver solution

Robot applications are closely tied to motor driving, and ROHM's motor driver solution includes essential components for motor driving, such as MOSFET for wheel systems related to autonomous robot movement and controllers for various motion controls. These solutions are characterized by their compact size and high efficiency.

For battery-powered autonomous mobile robots, the mainstream power supply solution is DC 24-72V, while industrial robots primarily use DC 48V power. With the advancement of technologies such as AI, sensing (image recognition), and increased driving capability, robot functionalities continue to upgrade, resulting in higher power consumption. In this context, power conversion technology has become an increasingly prominent issue and a pressing challenge to address. Power conversion efficiency in the field of AI robotics technology has become a critically important factor.

ROHM has developed a solution by combining external Nch MOSFET, motor drivers suitable for DC 48V applications, and the high-side/low-side gate driver IC "BD2320UEFJ-LA" with a voltage rating of 100V. This solution is further enhanced by utilizing N-channel MOSFETs from the "RS6xxxxBx/RH6xxxxBx series", which feature a copper clip structure, compact packaging, and low power losses. By reducing switch and conduction losses, this solution greatly contributes to improving the operational efficiency of application products.

Figure 5: Illustration of ROHM's gate driver IC and Nch MOSFET packaging

For example, when comparing power efficiency on an evaluation board for industrial equipment, the RS6xxxxBx/RH6xxxxBx series demonstrates a peak power efficiency of up to 95.01% within the output voltage range during steady-state operation.

Figure 6: Efficiency comparison between RS6xxxxBx/RH6xxxxBx series and conventional products

Conclusion

Robots combined with 3D perception LiDAR solutions and integrating AI technology will propel robot applications into a new era. ROHM has developed the ROHM LiDAR solution specifically for advanced autonomous mobility and various motion control applications in the field of AI robotics. This solution combines laser diodes with industry-leading ultra-small light source size, GaN HEMT (EcoGaNTM) for high-speed laser diode driving, and gate driver IC for GaN HEMT driving, enabling the construction of systems capable of obtaining higher-precision point cloud data.

In addition, ROHM offers motor drive solutions that enhance the efficiency of application products by combining high-side and low-side gate driver IC suitable for motor driving with Nch MOSFET featuring a compact package size, low power loss, and employing a copper clip structure. Furthermore, ROHM provides related reference designs and a free simulation tool called "ROHM Solution Simulator" on its official website. By effectively utilizing these resources, users can reduce design and evaluation time, significantly shortening the product introduction cycle to the market. These solutions make ROHM an ideal choice for relevant robot applications.

 

Neue Beiträge

Leider ergab Ihre Suche kein Ergebnis

Aktuelles über Elektronikkomponenten­

Wir haben unsere Datenschutzbestimmungen aktualisiert. Bitte nehmen Sie sich einen Moment Zeit, diese Änderungen zu überprüfen. Mit einem Klick auf "Ich stimme zu", stimmen Sie den Datenschutz- und Nutzungsbedingungen von Arrow Electronics zu.

Wir verwenden Cookies, um den Anwendernutzen zu vergrößern und unsere Webseite zu optimieren. Mehr über Cookies und wie man sie abschaltet finden Sie hier. Cookies und tracking Technologien können für Marketingzwecke verwendet werden.
Durch Klicken von „RICHTLINIEN AKZEPTIEREN“ stimmen Sie der Verwendung von Cookies auf Ihrem Endgerät und der Verwendung von tracking Technologien zu. Klicken Sie auf „MEHR INFORMATIONEN“ unten für mehr Informationen und Anleitungen wie man Cookies und tracking Technologien abschaltet. Das Akzeptieren von Cookies und tracking Technologien ist zwar freiwillig, das Blockieren kann aber eine korrekte Ausführung unserer Website verhindern, und bestimmte Werbung könnte für Sie weniger relevant sein.
Ihr Datenschutz ist uns wichtig. Lesen Sie mehr über unsere Datenschutzrichtlinien hier.