Driving the autonomous revolution of the logistics sector, a cutting-edge Vision Only Robot Solution has been developed by Geek+.
Driving the smart revolution of the logistics sector, cutting-edge technology is being developed to equip autonomous robots and vehicles to perform more tasks, more efficiently than ever before.
Claimed as the “world’s first vision only robot solution”, Geek+’s Vision Only Robot Solution integrates with a Visual Navigation Module from Intel. The solution includes depth vision perception through the use of an Intel RealSense camera and offers algorithm innovation with V-SLAM positioning, composite detection networks and robot following.
The combination of technologies provides robots with highly accurate positioning, navigation and obstacle avoidance capabilities, meaning logistics solutions providers are able to complete complex and diverse tasks whilst improving both accuracy and efficiency.
Depth Vision Perception
Integrating with the company’s Visual Navigation Module, Intel’s RealSense camera provides support for vision-based AI solutions – accelerating the machine learning process and shortening the deployment cycle of automation when paired with a dedicated visual processor – whilst the system’s all-in-one design results in lower power consumption.
Depth data provided by the RealSense camera allows the robot to observe, understand, interact with and even learn from its environment. Meanwhile, the robot’s Visual Navigation Module also includes a Robotic Vision Hub, which provides computational support for V-SLAM positioning, composite detection networks and robot following.
“The Vision Only Robot Solution effectively leverages the depth vision perception of the Intel RealSense camera,” said Solomon Lee, VP of Product at Geek+. “Together with deep algorithmic innovations from both sides, it results in a boost in business growth and efficiency for customers, driving the digital and intelligent upgrade of smart logistics.”
V-SLAM positioning algorithm: Fusing multi-sensor data and various visual feature elements to generate composite maps, V-SLAM positioning delivers highly reliable and precise positioning in complex dynamic environments.
Composite Detection Network: Using both traditional object detection and validation networks, the composite detection network processes detection data from multiple dimensions, enhancing accuracy and reducing false detections.
Robot Following: By integrating personnel detection, re-identification and visual target tracking, the robot is able to accurately determine its position in relation to the target. Meanwhile, Geek+’s standalone RoboGo system includes a local planning algorithm to enable autonomous obstacle avoidance.
Enabling agile autonomous logistics to strengthen supply chains
“Highly accurate and consistent depth vision is critical for AMR (autonomous mobile robots) to achieve environmental perception, significantly influencing its performance in positioning, navigation and obstacle avoidance,” said Mark Yahiro, VP of Corporate Strategy and Ventures and the General Manager of Intel’s RealSense Business Unit.
With the introduction of smart logistics, warehousing service providers are expected to adopt autonomous robots and systems, helping customers build intelligent and agile digital supply chains.
