ams OSRAM’s Belago 1.1 Lights Way for Luxonis‘ New 3D Vision System for Robots

• ams OSRAM’s Belago 1.1 Dot Projector is featured in Luxonis’ new OAK-D Pro spatial Artificial Intelligence (AI) cameras.
• Belago 1.1 Dot Projector combines an infrared Vertical Cavity Surface Emitting Laser (VCSEL) with advanced optics to enable 3D sensing solutions.
• Many of today’s robots, automatic guided vehicles (AGVs) and drones rely on complete 3D sensing technology that uses infrared light to move autonomously in a facility.

ams OSRAM (SIX: AMS), a global leader in optical solutions, announced today a new partnership with Luxonis, a developer of embedded artificial intelligence and computer vision technology. Luxonis provides 3D solutions for automatic guided vehicles (AGVs), robots, drones and more that help create high-quality 3D maps for applications such as object detection and obstacle avoidance. A key component of these solutions are infrared light sources, such as ams OSRAM’s Belago 1.1 Dot Projector. The Belago 1.1 combines an infrared VCSEL chip with special optics and a robust package that is ideally suited for Active Stereo Vision (ASV).

With ASV, two infrared cameras read the pattern Belago 1.1 projects on a defined field-of-view. By comparing the images of both cameras, depth is calculated, and a 3D map is created. Obstacles are detected on the 3D map prompting the robot to take action, such as stopping or going around the obstacle.

Luxonis chose ams OSRAM’s Belago 1.1 for its recent OAK-D Pro spatial AI camera. The OAK-D Pro provides fast and accurate object detection through AI. Its open-source software allows for fast and flexible integration into a wide variety of application environments. Stereo cameras, featuring the Belago 1.1 Dot Projector, offer accurate depth perception for any scene and lighting condition for object tracking in real-time.

“Active Stereo is incredibly powerful and useful in robotics – but it was a missing piece in the OAK ecosystem until now – primarily because we wanted a solution with rock-solid eye safety. Given the open nature of our ecosystem, it was pivotal that we be eye safe. We were ecstatic when we found Belago 1.1 because its architecture allows simple and robust eye-safety, which enables Luxonis and our customers to autonomously integrate active stereo depth safely and confidently, unlocking untold applicability,” said Brandon Gilles, CEO at Luxonis.

The Belago 1.1 Dot Projector features an infrared pattern projector for increased depth performance. The product includes 5,000 individual light dots, which are generated by combining a VCSEL with a Micro Lens Array (MLA) optical system.

“The Belago 1.1 provides very stable, focus-free dot projection for increased texture on the scene and ASV depth performance. Applied in Luxonis’ latest ASV camera, customers have access to a turnkey 3D solution for quick adoption in a wide range of applications,” said Joerg Wertli, Senior Marketing Manager at ams OSRAM.

The compact dimensions of 4.2 mm x 3.6 mm x 3.3 mm allow customers to develop space-saving system designs. The Belago 1.1 emits light with a wavelength of 940 nanometers for operation independent of ambient light conditions. The robot’s path is reliably detected at a distance of one to three meters for collision avoidance. The Luxonis OAK-D Pro also uses ams OSRAM’s SFH 4725AS A01 infrared LED for flood illumination for no-light compute vision and perception. Visit our website for more information about the Belago 1.1 Dot Projector. For more information about Luxonis’ solutions, please visit luxonis.com.

About Luxonis

Luxonis provides holistic perception solutions for robotics and robotic automation. We enable our customers to focus on their business-value problems instead of the underlying robotics perceptions problems. To do so, we build and continually update an ecosystem of cloud management, AI training, software, firmware and hardware specifically focused on making robotic perception easy, called DepthAI.

Our suite of OAK-D cameras embed performant spatial sensing, neural inference (e.g. object detection, semantic segmentation, etc.), and computer vision functionality (motion estimation, optical flow, feature extraction & tracking, etc.) – with all computation done on camera – offloading the entire robotics perception problem to the camera itself. This leaves the robot CPU free for business-value problems, rather than being tied up running depth, AI, and CV vision pipelines.

And our series of OAK-SoM allow embedding all of this power directly into custom products at low technical risk and with specific optics, field of view, and image sensors/placement to suit application-specific needs. The OAK-D series and OAK-SoM are now the go-to solution for robotics perception.