The multitude of inexpensive sensors available on today’s robots enable a higher level of intelligence through sensor fusion – via sophisticated computer algorithms that extract and estimate new information.
The open source Kauai Labs Sensor Fusion Framework (SF2), a key component of the Build Better Robots® platform, is comprised of software libraries and tools to quickly and easily fuse data from various sensors, enabling several key new features for autonomous and driver-assisted navigation:
|Fusion-enabled Capability||Component Sensors||SF2 Release|
|Video Processing Latency Correction||Video Camera, IMU||Current release|
|IMU Odometry (Dead-reckoning)||Quadrature Encoders, IMU||Release 2: est. Fall, 2018|
|Robot Localization||Video Camera, LIDAR, Quadrature Encoders, IMU||Release 3: est. Summer, 2019|
NOTE: Development has been delayed due to the release of the VMX-pi product and integration with ROS. Release 2 which will implement IMU Odometry fusion algorithms. Release 3 is also planned to implement Robot Localization.
Designed to integrate easily into FRC and FTC Robot Control Systems, SF2:
- acquires data streams from multiple sensors
- synchronizes sensor data streams
- interpolates data from low-sample rate sensors
- fuses sensor data streams using state-of-the-art algorithms
- includes tools for debugging, data visualization and offline data analysis
SF2 works seamlessly with Kauai Labs Sensors (navX-MXP, navX-Micro) and controllers (VMX-pi) and supports multiple robot platforms including FRC robotics, FTC robotics and Linux/Windows-based robot control systems. SF2 is an open framework and can work with any third-party sensors which implement the SF2 sensor data source interface.
SF2 includes tutorials and examples with source code in several popular programming languages to streamline the integration of advanced sensor fusion into a robot control system.