I am currently working on a LiDAR-based autonomous spraying system for orchards. The system will be semi-autonomous, where a human drives the tractor manually, and the attached spraying system operates autonomously.
Current Prototype Setup:
Mobile Platform: Replaced the tractor with a mobile robot for testing.
Sprayer: Eight nozzles.
LiDAR: Slamtec RPLiDAR A2M12 360 Degree Laser Scanner Kit – 12M Range.
Processing Unit: Asus laptop with 8GB RAM, Intel i7-10870H CPU @ 2.20GHz, running Ubuntu 20.04 with ROS Noetic. Arduino Mega.
Data Flow: LiDAR data is sent to ROS via USB, preprocessed in Python (with C libraries), and zones of density are calculated. This data is then sent to an Arduino Mega via serial communication to control the sprayer valves through relays.
Results: Successful experiments with this setup.
Objective:
To reduce computational costs and invest in a higher-quality LiDAR, I am transitioning to an embedded circuit with low-level programming, aiming for a budget-friendly development board under Rs.6000 (~$72). I am relatively new to embedded systems, typically using Arduino, Jetson, or Raspberry Pi for projects.
New Setup Requirements:
LiDAR: Richbeam LakiBeam 1 – 270 Degree Industrial-grade Single-line 2D ToF LiDAR (25m, 12V, Ethernet UDP, IP65).
Specifications:
- Range Principle: TOF
- Detection Distance: 0-25m
- Data Sample Rate: 36kHz;28.8kHz
- Range Accuracy: ±2cm
- Power Consumption: ≤2W
- Horizontal FOV: 270°
- Horizontal Resolution & Rotation Frequency: 0.1°;0.25°
Additional Sensor: Optical encoder to continuously monitor the speed of the tractor.
Current Considerations:
I am considering the Teensy 4.1 or STM32 Cortex M series boards but am open to other suggestions that would fit within the budget and meet the performance requirements.
Question:
Which embedded board would be the most suitable for this application, considering the need to handle LiDAR data processing, control the sprayer valves, and monitor the tractor speed?
Thank you for your suggestions and insights!
