A Heavy Duty Smart Robot – In Progress
This is a relatively large and heavy duty robotic system that will eventually feature ROS and lidar-based SLAM navigation. Featuring four brushless outrunner hub motors which are commonly found on Segways and hoverboards, this robot can theoretically carry hundreds of pounds at human walking speed.
The v-slot 2020 aluminum frame is durable and modular, capable of supporting additional subsystems such as robotic arms, sensor racks, and actuators. This robot will eventually be powered by an Nvidia Jetson Nano single board computer running Robot Operating System to be capable of lidar SLAM navigation as well as specialized computer vision programs such as human detection and object avoidance.
It features a Lidar 360 degree rangefinder and has motor feedback. I took advantage of 3D printing technology for various components.
- Mount electronic components to frame
- Control the velocity of the robot using an RC controller
- Devise method of displaying robot intent to human operators (Via LEDs, buzzers, monitors)
- Install and configure ROS
- Program Simultaneous Localization and Mapping (SLAM) with Lidar
- Attempt Autonomous Navigation with Obstacle Avoidance
- Begin using computer vision to identify objects and follow humans
Possible Project Futures
- Integration with GPT3 and ChatGPT through OpenAI’s API to control the robot via artificial intelligence interpretation of human verbal intent. (Ex: human verbally asks questions to robot or tells it to go someplace, follow them, or perform other tasks)