Assessments and Grading

Assessment Philosophy

This course evaluates your ability to build functional robotic systems, not just understand theory. Assessments focus on:

Working implementations - Code that runs and demonstrates key concepts
System integration - Combining multiple ROS 2 components effectively
Problem-solving - Debugging and adapting to challenges
Documentation - Clear explanations of your design decisions

Assessment Structure

Module	Assessment Type	Weight	Due Week
Weeks 1-5	ROS 2 Mini-Project	20%	Week 5
Weeks 6-7	Simulation Project	15%	Week 7
Weeks 8-10	NVIDIA Isaac Project	20%	Week 10
Weeks 11-12	Humanoid Control Project	20%	Week 12
Week 13	Capstone Project	25%	Week 13

Total: 100%

Week 5 Assessment: ROS 2 Mini-Project

Project Requirements

Objective: Build a multi-node ROS 2 system that demonstrates core robotics concepts.

Deliverables:

Source Code - Complete ROS 2 package with:
- At least 3 nodes (sensor, processor, actuator)
- Topic-based communication between nodes
- One service for configuration/commands
- Launch file to start all nodes
Documentation - README.md with:
- System architecture diagram
- Node descriptions and interfaces
- Build and run instructions
- Sample output/screenshots
Video Demo - 2-3 minute screencast showing:
- System startup
- Data flow visualization (rqt_graph)
- Example use case execution

Grading Rubric (100 points)

Criterion	Excellent (A)	Good (B)	Satisfactory (C)	Needs Work (D/F)
Functionality (40pts)	System works reliably, handles edge cases	Works for main use cases, minor issues	Basic functionality present, some bugs	Doesn't run or major issues
Architecture (25pts)	Well-designed, modular, follows ROS 2 best practices	Good structure, mostly follows conventions	Basic structure, some design issues	Poor structure, tightly coupled
Documentation (20pts)	Comprehensive, clear diagrams, easy to replicate	Good explanations, some details missing	Basic documentation, requires clarification	Missing key information
Code Quality (15pts)	Clean, commented, proper error handling	Readable, basic comments	Functional but messy	Hard to understand, no comments

Submission Guidelines

Format: Git repository (GitHub/GitLab) with public access
Deadline: End of Week 5 (Friday 11:59 PM)
Submit: Repository URL via course platform
Late Policy: -10% per day, max 3 days

Week 7 Assessment: Simulation Project

Project Requirements

Objective: Create a simulated robot environment in Gazebo with autonomous behavior.

Deliverables:

Gazebo World - Custom simulation environment with:
- At least 3 obstacles/objects
- Robot model with sensors (camera, lidar, or IMU)
- Physics-based interactions
Autonomous Behavior - Robot that performs:
- SLAM (Simultaneous Localization and Mapping)
- Path planning to goal locations
- Obstacle avoidance
Analysis Report - PDF document (3-5 pages) with:
- System design and sensor choices
- Algorithm explanations (SLAM, navigation)
- Performance metrics (success rate, time to goal)
- Challenges faced and solutions

Grading Rubric (100 points)

Criterion	Excellent (A)	Good (B)	Satisfactory (C)	Needs Work (D/F)
Simulation Setup (25pts)	Complex environment, realistic physics	Good environment, basic physics	Simple environment, works	Environment doesn't load
Autonomous Navigation (35pts)	Reliable navigation, handles dynamic obstacles	Navigates successfully most times	Basic navigation, some failures	Doesn't navigate or frequent crashes
SLAM Quality (20pts)	Accurate map, handles loop closure	Good map with minor errors	Basic map, significant drift	No map or unusable
Report (20pts)	Thorough analysis, clear metrics	Good explanations, results shown	Basic documentation	Missing critical information

Submission Guidelines

Format: Git repository + PDF report
Deadline: End of Week 7 (Friday 11:59 PM)
Include: Recorded video (3-5 min) of robot navigating

Week 10 Assessment: NVIDIA Isaac Sim Project

Project Requirements

Objective: Develop a perception-based manipulation task using NVIDIA Isaac Sim.

Deliverables:

Isaac Sim Scene - Simulated environment with:
- Robotic manipulator (UR5, Franka, or similar)
- Objects to manipulate (at least 3)
- RGB-D camera for perception
Perception Pipeline - Computer vision system that:
- Detects objects using deep learning (YOLO, PointNet, etc.)
- Estimates 6D poses
- Handles occlusion and lighting variations
Manipulation Task - Robot performs:
- Pick-and-place operations
- Collision-free trajectory planning (MoveIt2)
- Success rate over 70% over 10 trials
Technical Report - PDF (5-7 pages) covering:
- Perception algorithm details
- Motion planning approach
- Experimental results with failure analysis

Grading Rubric (100 points)

Criterion	Excellent (A)	Good (B)	Satisfactory (C)	Needs Work (D/F)
Perception (30pts)	Robust detection, accurate pose estimation	Good detection, minor pose errors	Basic detection, significant errors	Detection fails frequently
Manipulation (30pts)	Smooth motions, greater than 80% success rate	Works well, 60-80% success	Basic function, less than 60% success	Doesn't complete tasks
Integration (20pts)	Seamless perception-to-motion pipeline	Good integration, minor delays	Basic integration, noticeable lag	Poor integration or crashes
Report (20pts)	In-depth analysis, clear metrics, failure discussion	Good documentation, results shown	Basic report, minimal analysis	Incomplete or unclear

Submission Guidelines

Format: Git repository + PDF report + demo video
Deadline: End of Week 10 (Friday 11:59 PM)
Video: 5-minute demonstration of perception + manipulation pipeline

Week 12 Assessment: Humanoid Control Project

Project Requirements

Objective: Implement locomotion or manipulation on a humanoid robot (sim or real).

Deliverables:

URDF Model - Complete humanoid description with:
- At least 12 DOF (arms + legs)
- Proper mass and inertia properties
- Visual and collision geometry
Control System - Implement one of:
- Option A: Bipedal walking (at least 5 consecutive steps)
- Option B: Upper-body manipulation (reach, grasp, place)
- Option C: Whole-body control (locomotion + manipulation)
Stability Analysis - Demonstrate:
- ZMP (Zero Moment Point) tracking
- Balance recovery from small pushes
- Joint torque limits respected
Project Report - PDF (6-8 pages) with:
- Kinematic and dynamic model description
- Control algorithm details (PID, MPC, RL, etc.)
- Experimental results with graphs
- Video frames showing key phases

Grading Rubric (100 points)

Criterion	Excellent (A)	Good (B)	Satisfactory (C)	Needs Work (D/F)
Model Quality (20pts)	Accurate URDF, realistic dynamics	Good model, minor inaccuracies	Basic model, some issues	Model doesn't load or unstable
Control Performance (35pts)	Robust, smooth motions, handles disturbances	Good performance, occasional issues	Basic function, frequent failures	Doesn't achieve task
Stability (25pts)	Excellent balance, ZMP within support polygon	Good stability, minor violations	Marginal stability	Falls or unstable
Report (20pts)	Comprehensive analysis, clear methodology	Good documentation, results explained	Basic report, minimal detail	Incomplete or unclear

Submission Guidelines

Format: Git repository + PDF report + video demo
Deadline: End of Week 12 (Friday 11:59 PM)
Video: 3-5 minutes showing task execution and stability tests

Week 13: Capstone Project

See Capstone Project Details for complete requirements and rubric.

Summary: Design and implement an autonomous humanoid robot that combines:

Perception (vision, depth sensing)
Navigation (SLAM, path planning)
Manipulation (grasping, placing)
Locomotion (walking, balance control)

Weight: 25% of final grade
Team Size: 1-3 students
Duration: Weeks 13 (intensive focus)

Grading Scale

Letter Grade	Percentage	Description
A	90-100%	Exceptional work, exceeds expectations
B	80-89%	Good work, meets all requirements
C	70-79%	Satisfactory work, meets basic requirements
D	60-69%	Below expectations, significant issues
F	Below 60%	Unsatisfactory, does not meet minimum standards

General Submission Guidelines

Repository Structure

project-name/
├── README.md               # Project overview and instructions
├── package.xml             # ROS 2 package metadata
├── CMakeLists.txt          # Build configuration (if C++)
├── setup.py                # Setup file (if Python)
├── launch/                 # Launch files
│   └── demo.launch.py
├── src/                    # Source code
│   ├── node1.py
│   └── node2.py
├── config/                 # Configuration files (YAML, JSON)
├── worlds/                 # Gazebo world files (if applicable)
├── urdf/                   # Robot descriptions
├── docs/                   # Additional documentation
│   ├── architecture.md
│   └── diagrams/
└── media/                  # Screenshots, videos
    └── demo.mp4

Video Requirements

Format: MP4, MOV, or WEBM
Resolution: Minimum 720p
Duration: As specified per assessment
Audio: Include narration explaining what's happening
Host: YouTube (unlisted), Google Drive, or Vimeo

Code Standards

Python: Follow PEP 8 style guide
C++: Follow ROS 2 C++ style guide
Comments: Explain non-obvious logic
Error Handling: Include try/except blocks for critical operations
Logging: Use ROS 2 logging (get_logger().info(), etc.)

Documentation Standards

README.md: Must include:
- Project description and goals
- Prerequisites and dependencies
- Build instructions
- Run instructions with example commands
- Troubleshooting common issues
Code comments: Docstrings for all functions/classes
Diagrams: Use draw.io, Lucidchart, or similar
Report PDFs: Use proper headings, figures with captions, references

Academic Integrity

Collaboration: Permitted for learning, but all submitted code must be your own
External Code: You may use open-source libraries, but cite all sources
AI Tools: GitHub Copilot and ChatGPT are permitted as coding assistants, but you must understand all code you submit
Plagiarism: Copying code from classmates or online sources without attribution results in automatic zero

Late Submission Policy

First 24 hours: -10% penalty
24-48 hours: -20% penalty
48-72 hours: -30% penalty
Beyond 72 hours: Zero credit unless prior arrangement with instructor

Extensions: Available for medical emergencies or documented circumstances. Contact instructor at least 48 hours before deadline.

Getting Help

Office Hours: See course syllabus for schedule
Discord/Slack: Use course channel for questions (response within 24 hours)
Email: For private matters only (allow 48 hours for response)
Tutoring: Peer tutoring available, schedule through course platform

Assessment Tips

For Success:

Start early - Don't wait until the last week
Test incrementally - Verify each component before integration
Document as you go - Don't leave documentation until the end
Use version control - Commit frequently with meaningful messages
Ask questions - Use office hours and discussion forums
Record videos early - Technical issues can delay final submission

Common Mistakes to Avoid:

❌ Overly complex designs that are hard to debug
❌ Not testing on clean system (missing dependencies)
❌ Poor git practices (single massive commit)
❌ Waiting until deadline to record video
❌ Not reading assignment requirements carefully
❌ Hardcoding paths specific to your machine

Frequently Asked Questions

Q: Can I use a different simulator than Gazebo?
A: For Week 7, yes (Unity, Webots, PyBullet). For Week 10, NVIDIA Isaac Sim is required.

Q: What if my robot keeps falling in simulation?
A: Start with static balance tests, then add motion. Use lower gains for PID control. Check mass/inertia properties.

Q: Can I work in a team?
A: Weeks 1-12 are individual. Capstone project allows teams of 1-3.

Q: What hardware do I need?
A: See Hardware Requirements. Minimum: laptop with 8GB RAM for simulation.

Q: How do I submit if my repository is private?
A: Either make it public or add instructor as collaborator.

Resources

For additional support, visit the course discussion forum or attend office hours.

Assessment Philosophy​

Assessment Structure​

Week 5 Assessment: ROS 2 Mini-Project​

Project Requirements​

Suggested Topics​

Grading Rubric (100 points)​

Submission Guidelines​

Week 7 Assessment: Simulation Project​

Project Requirements​

Grading Rubric (100 points)​

Submission Guidelines​

Week 10 Assessment: NVIDIA Isaac Sim Project​

Project Requirements​

Grading Rubric (100 points)​

Submission Guidelines​

Week 12 Assessment: Humanoid Control Project​

Project Requirements​

Grading Rubric (100 points)​

Submission Guidelines​

Week 13: Capstone Project​

Grading Scale​

General Submission Guidelines​

Repository Structure​

Video Requirements​

Code Standards​

Documentation Standards​

Academic Integrity​

Late Submission Policy​

Getting Help​

Assessment Tips​

For Success:​

Common Mistakes to Avoid:​

Frequently Asked Questions​

Resources​

Assessment Philosophy

Assessment Structure

Week 5 Assessment: ROS 2 Mini-Project

Project Requirements

Suggested Topics

Grading Rubric (100 points)

Submission Guidelines

Week 7 Assessment: Simulation Project

Project Requirements

Grading Rubric (100 points)

Submission Guidelines

Week 10 Assessment: NVIDIA Isaac Sim Project

Project Requirements

Grading Rubric (100 points)

Submission Guidelines

Week 12 Assessment: Humanoid Control Project

Project Requirements

Grading Rubric (100 points)

Submission Guidelines

Week 13: Capstone Project

Grading Scale

General Submission Guidelines

Repository Structure

Video Requirements

Code Standards

Documentation Standards

Academic Integrity

Late Submission Policy

Getting Help

Assessment Tips

For Success:

Common Mistakes to Avoid:

Frequently Asked Questions

Resources