In the fast-evolving world of Autonomous Mobile Robots (AMRs), innovation isn’t just about groundbreaking algorithms or cutting-edge sensors – sometimes, it’s about a fresh perspective on how to integrate existing, powerful technologies. We recently encountered a fascinating design challenge from a customer aiming to build next-generation AMRs with a particularly slim form factor. Their goal: advanced navigation, robust collision avoidance, and precise SLAM (Simultaneous Localization and Mapping), all within a confined space.
The solution? A clever orientation of a high-performance LiDAR, paired with the unparalleled AI processing power of NVIDIA’s Orin series GPUs.
AMRs are increasingly being deployed in environments where space is a premium – think crowded warehouses, narrow factory aisles, or public spaces. This often necessitates a low-profile design, making the integration of critical sensors like LiDAR a significant hurdle. Traditional LiDAR units, while powerful, can sometimes be too tall or bulky for these sleek designs.
Our customer needed a LiDAR that could deliver a comprehensive 3D point cloud for detailed environmental perception but couldn’t accommodate a standard, upright mounting.
Enter the Hesai JT128 LiDAR, distributed by CRATUS Technology. This isn’t just any LiDAR; it’s a 128-channel powerhouse designed specifically for robotics and industrial applications. While its impressive technical specifications – high resolution, a wide 360° (H) x 187° (V) field of view, and an integrated IMU – are well-known, it was its physical dimensions that sparked an “aha!” moment.
The JT128 boasts an approximately 50 mm diameter and 68 mm height. While 68 mm might still be too tall for some slim applications, we realized that by mounting the JT128 sideways, its roughly 50 mm dimension (the side profile when oriented horizontally) perfectly fit the customer’s stringent form factor requirements.
This seemingly simple reorientation opened up a world of possibilities, allowing the customer to achieve their slim design without compromising on sensor performance. The JT128’s hyper-hemispherical FOV ensures no blind spots, providing robust environmental perception even in this novel mounting configuration. Its integrated IMU is crucial for accurate pose estimation and robust SLAM, especially when navigating complex paths.
A powerful sensor demands equally powerful processing. This is where the NVIDIA Orin series GPUs, also distributed by CRATUS Technology, come into play. The Orin platform (including Orin Nano, Orin NX, and AGX Orin) provides the industry-leading AI compute power necessary to crunch the massive amounts of data generated by the JT128’s high-resolution point cloud.
With an Orin GPU on board, the AMR can:
The seamless integration of the JT128’s point cloud and IMU data with the Orin’s processing prowess creates a formidable perception and navigation system.
This solution highlights the synergy between cutting-edge sensor technology and powerful AI computing. Partnering with CRATUS Technology, distributing both HESAI LiDARs and NVIDIA Orin GPUs, means customers can access a complete, integrated solution from a trusted source for robotics and warehouse automation.
The excitement from our customer upon seeing this “sideways” solution was palpable. It’s a testament to the idea that sometimes, the most elegant solutions are found by simply looking at a problem from a different angle – literally!
The right combination of hardware and an innovative approach can unlock incredible potential. If you’re facing similar design constraints in your robotics or autonomous vehicle projects, don’t hesitate to think outside the box… or in this case, rotate the box! Better yet, contact CRATUS Technology and we will brainstorm together.
