Skip to main content

Assessments

This chapter contains all module assessments and evaluation rubrics. Each assessment tests the skills learned in its corresponding module.

Module 1: ROS 2 Fundamentals​

Project: Mobile Robot Navigation System​

Objective: Build a complete ROS 2 application for a simulated mobile robot

Requirements:

  1. Teleoperation node (keyboard → /cmd_vel)
  2. Sensor fusion node (camera + odometry)
  3. Obstacle avoidance node (laser scan → reactive control)
  4. State publisher (robot pose → /tf)
  5. Launch file (start all nodes)
  6. ROS bag (record 1 minute of data)

Rubric (100 points):

  • Functionality (40 pts): Robot navigates 10m without collisions
  • Code quality (20 pts): Proper naming, structure, comments
  • ROS 2 best practices (20 pts): QoS, logging, parameters
  • Launch file (10 pts): Single-command startup
  • Documentation (10 pts): README with setup instructions

Deliverables:

  • Source code (GitHub repo)
  • Demo video (2-3 min)
  • README

Due: End of Week 5

Module 2: Robot Simulation​

Project: Mobile Manipulator in Gazebo​

Objective: Create a complete Gazebo simulation of a mobile manipulator

Requirements:

  1. Differential drive mobile base with odometry
  2. 5-DOF manipulator arm
  3. RGB-D camera + 2D LiDAR
  4. Grasping plugin for object interaction
  5. Warehouse environment with obstacles
  6. ROS 2 control interface

Rubric (100 points):

  • URDF model (25 pts): Accurate kinematics and inertias
  • Sensors (20 pts): Realistic noise models
  • Gazebo world (15 pts): Interesting environment
  • ROS 2 integration (20 pts): Topics for all sensors
  • Grasping (10 pts): Pick up objects reliably
  • Documentation (10 pts): URDF explained

Deliverables:

  • URDF file + meshes
  • Gazebo world file
  • Launch files
  • Demo video

Due: End of Week 7

Module 3: NVIDIA Isaac Platform​

Project: RL-Based Manipulation Policy​

Objective: Train a reinforcement learning policy for robot manipulation

Requirements:

  1. Task: Pick and place objects from randomized positions
  2. Environment: Isaac Sim with Franka Panda arm
  3. Algorithm: PPO or SAC
  4. Success metric: >80% on unseen configurations
  5. (Bonus) Transfer to real Franka arm

Rubric (100 points):

  • Training setup (20 pts): Reward function, hyperparameters
  • Policy performance (40 pts): Success rate on test set
  • Domain randomization (15 pts): Lighting, mass, friction
  • Code quality (15 pts): Clean, reproducible
  • Report (10 pts): Training curves, analysis

Deliverables:

  • Training code
  • Trained policy checkpoint
  • Evaluation results (CSV, plots)
  • Technical report (3-5 pages)

Due: End of Week 10

Module 4: Humanoid Development​

Project: Humanoid Locomotion Controller​

Objective: Implement a ZMP-based walking controller for a humanoid

Requirements:

  1. Walk 10 meters on flat ground
  2. Climb 3 stairs (20cm height)
  3. Recover from lateral push (50N force)
  4. Maintain balance while reaching with arms

Rubric (100 points):

  • Flat ground walking (25 pts): Smooth gait, no falls
  • Stair climbing (25 pts): Adaptive footstep planning
  • Push recovery (20 pts): Regain balance within 2 seconds
  • Arm reaching (15 pts): Coordinate with locomotion
  • Report (15 pts): Controller design, ZMP plots

Deliverables:

  • Controller code
  • Simulation world
  • Demo video (walking, stairs, push recovery)
  • Technical report

Due: End of Week 12

Final Capstone Project​

See Capstone Project for full details.

Grading Policy​

Assessment Weights:​

  • Module 1 (ROS 2): 15%
  • Module 2 (Simulation): 15%
  • Module 3 (Isaac): 20%
  • Module 4 (Humanoid): 20%
  • Capstone Project: 30%

Late Policy:​

  • 1-3 days late: -10%
  • 4-7 days late: -25%

  • 7 days late: -50%

Academic Integrity:​

  • You may discuss concepts with peers
  • All code must be your own
  • Cite external resources (papers, tutorials)
  • Plagiarism = automatic zero

Submission Instructions​

  1. Code: Push to GitHub (public or private repo)
  2. Video: Upload to YouTube (unlisted) or cloud storage
  3. Report: PDF via course management system
  4. Deadline: 11:59 PM on due date (your timezone)

Include this in your README.md:

# Module X Assessment

**Student**: Your Name
**Email**: your.email@example.com
**Date**: YYYY-MM-DD

## Demo Video
[Link to video]

## How to Run
\`\`\`bash
ros2 launch my_package assessment.launch.py
\`\`\`

## Results
- Metric 1: XX%
- Metric 2: YY seconds

## Reflections
What worked well, what was challenging, what you learned.

Need Help?​

  • Debugging: Use the AI chatbot (bottom-right corner)
  • Office Hours: Check course schedule
  • Discussion Forum: Post questions publicly
  • Email: For private issues only