Listen to this article

Autonomous navigation capabilities of mobile robots are highly dependent on the sensor stack. And with the use of mobile robots expanding across multiple industries, improved perception and navigation enables safer operation.

DreamVu, a Philadelphia startup founded in 2017, announced today its PAL and PAL Mini 3D vision systems are now available. Both PAL and PAL Mini provide high-resolution vision, high-accuracy depth, and color 3D point-clouds that can be used in a variety of systems, including autonomous mobile robots (AMRs), collaborative robots, disinfection robots, and domestic vacuum cleaning robots where low-latency, low-power, and high-reliability are critical.

Disinfection quickly became a popular application for robots during the COVID-19 pandemic. As part of our RoboBusiness Direct series, DreamVu chief revenue officer Mark Davidson will discuss how to make UV disinfection safe using omnidirectional 3D vision. Davidson will explain how omnidirectional 3D vision systems can be used to not only identify the presence of people, but to also identify how far away people are from the light based on accurate depth information. You can register for this RoboBusiness Direct talk, for free, here.

Both PAL and PAL Mini use only a single CMOS sensor and no moving parts. Both products provide visual intelligence that eliminates blind spots and combines full color video with precise depth mapping and detection in a single video stream.

PAL provides 360° horizontal field of view with 110° vertical field of view at up to 9 megapixels with a usable depth range of 10 meters and no minimum depth distance. An entire 3D occupancy map can be generated from every frame. PAL can enable AMRs to perform autonomous navigation, obstacle detection and avoidance, simultaneous localization and mapping (SLAM) and more.

Bengalaru, India-based Robotics Automation and Design (RAAD) Systems selected the PAL system for its AMP series AMRs, which started shipping to customers at the end of 2020. Davis Packiaraj, senior software engineer at RAAD Systems, said the AMP series will benefit from improved maneuverability thanks to the wide view provided by PAL. He said a single PAL system will replace two narrow field-of-view stereo cameras used in previous generations of the AMRs.

PAL Mini is a compact omnidirectional 3D vision system that delivers 360° horizontal field of view with a vertical field of view of 80° at up to 4 megapixels and a usable depth range of 5 meters. Due to its ultra-small form-factor, PAL Mini is optimized for navigation systems in smaller AMRs and domestic robots.

PAL Mini can also be used as a complementary sensor for applications such as people detection in disinfection robots, or wall, cliff, and stair tracking in commercial cleaning robots where maximizing cleaning area is paramount.

“The need for autonomous robots greatly accelerated in 2020 and we are excited to bring to market two game-changing vision systems,” said Davidson. “With PAL and PAL Mini, our customers can create autonomous robots that provide safe cleaning and disinfection, and support more efficient e-commerce operations.”

PAL is currently available with a USB interface and can be purchased for $1,499. An Ethernet version with on-board ARM processor and GPU is in development with prototypes also currently available for purchase. PAL Mini is available with a USB interface and can be purchased for $999.

Both PAL and PAL Mini come with a software development kit (SDK) with broad compatibility, including support for the Robot Operating System (ROS) and OpenCV. The products can also be complemented with DreamVu’s Vision Intelligence Software, which provides AI functions such as object recognition and classification, and people detection.

The Robot Report profiled DreamVu back in November 2019. You can read that profile here.