A ROS driver for TI mmWave radar sensors with sensor fusion, multi-radar support, and camera overlay capabilities.
TI mmWave ROS Package is a customized ROS driver for Texas Instruments mmWave radar sensors, providing enhanced data processing, multi-radar support, and camera overlay capabilities. It solves the problem of integrating radar sensors into robotic systems with real-time sensor fusion and flexible configuration.
Robotics researchers, autonomous systems developers, and engineers working with TI mmWave radar sensors who need ROS integration for perception and sensor fusion tasks.
Developers choose this package over TI's original driver because it adds critical features like Doppler data, multi-radar coordination, and camera overlay support, making it more suitable for complex real-world applications.
TI mmWave radar ROS driver (with sensor fusion and hybrid)
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
All radar parameters are exposed via ROS parameters, allowing dynamic configuration and monitoring, as shown in the ability to use `rosparam get` for values like max_doppler_vel.
Includes Doppler velocity data in custom ROS messages (/ti_mmwave/radar_scan), enabling richer perception for tracking moving targets, which is an improvement over TI's original driver.
Supports multiple radar devices working together with separate launch files and TF transforms, ideal for sensor arrays in complex robotic setups, as detailed in the multiple devices support section.
Enables sensor fusion by overlaying radar data with camera feeds, facilitated by launch files and calibration, enhancing perception for applications like autonomous navigation.
Requires manual steps like flashing devices with specific SDKs, setting serial port permissions, and ROS installation, which can be error-prone and time-consuming, as seen in the troubleshooting section.
Dependent on outdated ROS versions (Kinetic/Melodic) and specific TI mmWave models, with noted inconsistencies like AWR1843 requiring SDK 3.2.0.4 and different output formats, limiting flexibility.
Launching multiple radars requires separate launch files and manual adjustments to TF transforms, rather than automated coordination, increasing configuration overhead for scalable deployments.