Castalia Institute · Micro-Credentials

Field Robotics
Curriculum

Five stackable micro-credentials that take you from assembling a rover kit to leading autonomous field missions in the San Luis Valley.

AFR-100 Rover Assembly & AFR-200 Embedded Sensing & AFR-300 Computer Vision for AFR-400 Autonomous Navigation with AFR-500 Field Robotics Capstone
AFR-100 Foundation Phase 1 — Prototype

Rover Assembly & Teleoperation

Build a six-wheel rocker-bogie rover from a kit, learn basic Arduino programming, and operate the rover via FPV teleoperation. Covers mechanical assembly, motor control fundamentals, serial communication, and the SunFounder app for real-time driving.

40 Hours
$130 (kit) Materials
SunFounder GalaxyRVR Platform
None Prerequisites

Learning Outcomes

  • Assemble a rocker-bogie suspension rover from components
  • Write Arduino sketches for motor control and sensor reading
  • Configure ESP32 CAM for first-person video streaming
  • Operate a rover remotely using a mobile control app
  • Understand basic power budgeting and battery management

Topics Covered

  • Mechanical assembly & rocker-bogie kinematics
  • Arduino Uno R3 programming (C++)
  • DC motor drivers & PWM control
  • ESP32 CAM configuration & Wi-Fi streaming
  • Ultrasonic & infrared obstacle sensors
  • Solar panel charging basics
  • Serial communication protocols

Capstone Project

Complete a timed obstacle course with the GalaxyRVR under manual FPV control. Document the build process and produce a wiring diagram.

AFR-200 Intermediate Phase 2 — Sensor Integration

Embedded Sensing & Data Collection

Integrate environmental and navigation sensors into the rover platform. Program an ESP32-S3 to poll GPS, IMU, temperature, humidity, pressure, and soil moisture sensors, then log and transmit structured telemetry.

50 Hours
$80 (sensors) Materials
GalaxyRVR + sensor breakouts Platform
AFR-100 or equivalent Arduino experience Prerequisites

Learning Outcomes

  • Interface I2C, SPI, and UART sensors with an ESP32-S3
  • Fuse IMU data for orientation estimation
  • Log GPS tracks and environmental measurements to onboard storage
  • Transmit structured telemetry over Wi-Fi and LoRa
  • Calibrate sensors for field conditions

Topics Covered

  • ESP32-S3 development with ESP-IDF & Arduino
  • u-blox GNSS modules (ZED-F9P basics)
  • IMU fusion — Bosch BNO085 / ICM-20948
  • Environmental sensing — BME688 (T / H / P / VOC)
  • Soil moisture & analog sensor interfacing
  • Data logging to SD / SPI flash
  • Telemetry protocols — JSON over MQTT / LoRa
  • Sensor calibration & noise filtering

Capstone Project

Conduct a 30-minute autonomous data-collection run, logging GPS waypoints, environmental readings, and IMU orientation to onboard storage. Produce a map overlay of the sensor data.

AFR-300 Intermediate Phase 2–3 — Perception

Computer Vision for Field Robotics

Deploy computer vision on an edge compute platform. Covers stereo depth estimation, terrain classification with convolutional neural networks, and real-time obstacle detection — the perceptual foundation for autonomous navigation.

60 Hours
$350 (Jetson + camera) Materials
Jetson Orin Nano + OAK-D or monocular camera Platform
AFR-100, basic Python Prerequisites

Learning Outcomes

  • Configure a Jetson Orin Nano for vision workloads
  • Process stereo camera feeds for depth estimation
  • Train and deploy a terrain classification model
  • Implement real-time obstacle detection with bounding boxes
  • Evaluate model performance on desert terrain imagery

Topics Covered

  • NVIDIA Jetson setup — JetPack, CUDA, TensorRT
  • OpenCV fundamentals — image I/O, filtering, feature detection
  • Stereo vision — OAK-D depth pipeline (DepthAI)
  • Monocular depth estimation
  • CNN-based terrain classification (sand, gravel, vegetation, rock)
  • Object detection — YOLO / SSD on edge hardware
  • Dataset collection & annotation for field environments
  • Model optimization — quantization, pruning, TensorRT conversion

Capstone Project

Train a terrain classifier on San Luis Valley imagery and deploy it on the Jetson. Demonstrate real-time classification at 10+ FPS on a live camera feed, distinguishing at least four terrain types.

AFR-400 Advanced Phase 3 — Autonomy

Autonomous Navigation with ROS2

Build a full autonomous navigation stack using ROS2. Start in simulation with Gazebo, implement SLAM and path planning with Nav2, then transfer to the physical rover for waypoint missions in outdoor terrain.

80 Hours
$0 (simulation) / existing hardware Materials
ROS2 + Gazebo simulation, then rover hardware Platform
AFR-200, AFR-300 (or concurrent) Prerequisites

Learning Outcomes

  • Architect a ROS2 system with nodes, topics, and services
  • Simulate a rover in Gazebo with realistic terrain
  • Implement SLAM using RTAB-Map or ORB-SLAM
  • Configure Nav2 for path planning and obstacle avoidance
  • Execute autonomous waypoint missions on physical hardware

Topics Covered

  • ROS2 architecture — nodes, topics, services, actions
  • URDF / Xacro robot description
  • Gazebo simulation — world building, sensor plugins
  • tf2 coordinate transforms
  • SLAM — RTAB-Map (visual) & ORB-SLAM (feature-based)
  • Nav2 — costmaps, planners, controllers, behavior trees
  • Localization — AMCL, EKF sensor fusion
  • Sim-to-real transfer — tuning for outdoor conditions
  • Mission planning — waypoint sequencing & recovery behaviors

Capstone Project

Program the rover to autonomously navigate a 500m outdoor course with at least 5 waypoints, avoiding obstacles, and returning to the start position. Produce a SLAM-generated map of the traversed area.

AFR-500 Advanced Phase 4 — Field Deployment

Field Robotics Capstone

Plan and execute a multi-day autonomous field mission in the San Luis Valley. Covers environmental hardening, power management for long-duration operation, communications infrastructure, data pipelines, and scientific data analysis.

100 Hours
Included (shared fleet) Materials
Full AFR platform Platform
AFR-200, AFR-300, AFR-400 Prerequisites

Learning Outcomes

  • Prepare a rover for extended outdoor deployment (dust, wind, temperature)
  • Design a power budget with solar charging for multi-day operation
  • Deploy and manage field communications (Wi-Fi, LoRa, LTE)
  • Execute a multi-day autonomous survey mission
  • Build a data pipeline from rover to analysis workstation
  • Produce a scientific report from collected field data

Topics Covered

  • Environmental hardening — IP-rated enclosures, filtered cooling, thermal management
  • Power systems — LiFePO4 battery management, MPPT solar charging, power budgets
  • Field communications — mesh Wi-Fi, LoRa telemetry, LTE failover, Starlink base station
  • Mission planning — survey design, coverage algorithms, contingency handling
  • Data pipelines — onboard logging, wireless sync, cloud ingest
  • Geospatial analysis — GPS track processing, terrain model generation
  • Environmental data analysis — time-series visualization, anomaly detection
  • Field safety & logistics — site assessment, recovery procedures, regulatory compliance

Capstone Project

Lead a 48-hour autonomous survey mission in the San Luis Valley. The rover must navigate a pre-planned route, collect environmental and terrain data, manage its own power via solar charging, and transmit telemetry to a base station. Deliver a field report with maps, environmental analysis, and lessons learned.

Credential Structure

Individual Micro-Credentials

Each course awards a standalone micro-credential upon completion of the capstone project. Take any course independently if you meet the prerequisites.

Field Robotics Certificate

Complete all five micro-credentials (AFR-100 through AFR-500) to earn the Castalia Institute Field Robotics Certificate—330 hours of hands-on learning from assembly to autonomous field deployment.

330 total hours

Simulation Track

AFR-400 (Autonomous Navigation) can be completed entirely in simulation using ROS2 and Gazebo—no hardware required. Ideal for remote learners.