WHAT IS VISUAL SLAM?
VISUAL SLAM IS A TECHNOLOGY BASED ON COMPUTER VISION FOR PRECISE INDOOR LOCATION AND POSITIONING.
Visual SLAM refers to the complex process of calculating the position and orientation of a device with respect to its surroundings, while mapping the environment at the same time, using only visual inputs from a camera.
Accuware Dragonfly is an example of visual SLAM technology.
Introduction to Visual SLAM
Visual SLAM, also known as vSLAM, is a technology able to build a map of an unknown environment and perform location, simultaneously leveraging the partially built map, using just computer vision.
SLAM stands for “Simultaneous Localization and Mapping”. This means that the device performing SLAM is able to:
- Map the location, creating a 3D virtual map
- Locate itself inside the map
Visual SLAM uses only visual inputs to perform location and mapping, meaning that the only sensor required is a camera that has to be mounted on board of the device. No other external sensors are required.
Today, with the great improvements in automation and robotics, vSLAM is one of the most challenging open problems for developing autonomous robots and vehicles.
We have developed and created Accuware Dragonfly to solve this problem, and we are proud to offer a robust and reliable technology for precise indoor location of robots, drones, autonomous vehicles.
Dragonfly is a unique patented vSLAM technology that can work with monocular and stereo cameras, and is able to provide an accuracy up to 5 cm, without the use of LiDARs or of motion sensors.
Do you want to know more about Dragonfly?
How can visual SLAM be used and what are the applications?
Visual SLAM can be used in many ways, and its main scope is to provide precise location to autonomous devices, robots, drones, vehicles.
We work with different companies all around the world to address multiple requirements and projects with Dragonfly. This is a partial list of the typical use cases that can be addressed by Dragonfly:
- Provide location to robots and drones. Dragonfly comes with a direct ROS integration, upon request, becoming the first SLAM for ROS technology;
- Remotely monitor the location of devices: Dragonfly can be used to track forklifts inside a warehouse for example, to optimize the operations;
- Integrate autonomous navigation on board of Unmanned Ground or Aerial Vehicles (UGV, UAV);
- Control the movements of machines for collision avoidance.
DRONES POSITIONING AND NAVIGATION
Drones can be programmed to use Dragonfly as a location engine, and the location data can be used as a source for autonomous navigation. ROS nodes available upon request to integrate Dragonfly SLAM for ROS.
ROBOTICS AND AUTOMATION
Robots and automated devices can be tracked using Dragonfly, with an extremely high level of precision. ROS nodes available upon request to integrate Dragonfly SLAM for ROS.
It is possible to install Dragonfly on Forklifts, to monitor their movements inside warehouses and improve the operation management.
Dragonfly is a key component for indoor navigation. Where GPS is not precise enough, or inside GPS-denied environments, Dragonfly can provide the optimal level of precision for autonomous navigation.
Feel free to contact us to speak about your use case: we will be glad to help and suggest the best architecture.
How can vSLAM be integrated?
- X, Y, Z coordinates, expressed in meters from an origin.
- Absolute WGS-84 coordinates (latitude, longitude) + altitude in meters.
- Yaw, Pitch, Roll : these values indicate the rotation of the device on the three axes. You can read more about the principle behind the Aircraft principal axes on Wikipedia.
We are open to customize the API output upon request: our R&D team can tailor the payload to address different requirements (it is possible to add the speed indication, the deliver the output in other geographic protocols, to change the type of format…). Feel free to contact us if you have custom requests.
SLAM for ROS
Accuware provides ROS nodes for direct ROS integration so that Dragonfly can be seamlessly integrated on board of Robots and devices using ROS.