Your Hardware Tier
This course meets you where you are. Some students have laptops; others have gaming PCs with RTX GPUs; others have embedded Jetson boards. Rather than pretending everyone has the same hardware, we've designed the curriculum to work across hardware tiers.
This lesson helps you identify your tier and understand which learning paths are available to you.
The Four Hardware Tiers
Tier 1: Laptop or Cloud (Cloud-Based Learning)
Equipment: Laptop/desktop (any OS) with internet access
What you can do:
- Run ROS 2 in cloud (TheConstruct)
- Write Python code in cloud terminal
- Use MockROS (browser-based simulation)
- Run Pyodide (Python in browser)
- Access visualization tools via remote desktop
What you cannot do:
- Run local GPU-intensive simulation
- Deploy to edge hardware
- Access local USB sensors (sensors are virtual/mocked)
Timeline for Module 1: ✅ Full module possible (4-5 weeks) Timeline for Module 2: ⚠️ Partial (can use cloud Gazebo, slower iteration) Timeline for Module 3: ⚠️ Limited (Isaac cloud slower than local) Timeline for Module 4: ❌ Not without paid cloud resources
Cost: $0 (cloud ROS 2 has free tier) or $10-50/month for premium cloud access
Tier 2: RTX GPU Workstation (Local Development)
Equipment: Laptop/desktop with NVIDIA RTX GPU (RTX 3050 or better)
What you can do:
- Run ROS 2 locally (no cloud dependency)
- Run Gazebo locally (full physics simulation)
- Run Isaac Sim locally (fast iteration)
- Work offline
- Deploy code quickly (seconds, not minutes)
What you cannot do:
- Deploy to real robots (no Jetson/robot hardware)
- Test edge constraints (no ARM architecture)
- Use real sensors (only simulated)
Timeline for Module 1: ✅ Full module in 3 weeks (faster iteration) Timeline for Module 2: ✅ Full module locally Timeline for Module 3: ✅ Full module, fast Timeline for Module 4: ⚠️ Possible but slower than Tier 3/4
Cost: $300-1500 for GPU upgrade, or already have hardware
Tier 3: Jetson Edge Device (Edge Deployment)
Equipment:
- NVIDIA Jetson board (Orin Nano $200, Orin AGX $3,000+)
- Plus laptop from Tier 1 or 2 for development
- Optional: Sensors (IMU, camera)
What you can do:
- Run ROS 2 on embedded ARM hardware
- Deploy code to edge device (no cloud dependency)
- Connect real sensors (USB cameras, serial IMU)
- Experience real latency and resource constraints
- Build autonomous robots (with Tier 4 hardware optional)
What you cannot do:
- Run heavy GPU workloads locally (Isaac Sim must run on workstation, results stream to Jetson)
- Iterate as fast as Tier 2 (Jetson is slower than desktop GPU)
Timeline for Module 1: ✅ Full module Timeline for Module 2: ✅ Full module (development on workstation, deployment on Jetson) Timeline for Module 3: ✅ Full module Timeline for Module 4: ✅ Full module, realistic edge constraints
Cost: $200-3,000 for Jetson + $500-2,000 for sensors
Tier 4: Physical Robot (Real-World Testing)
Equipment:
- Tier 2 or 3 hardware for development
- Plus physical robot:
- Affordable options: Unitree Go2 ($1,500), Jetbot ($400)
- Expensive options: Figure humanoid ($150,000+), Tesla Bot (not yet available)
What you can do:
- Test code on real hardware
- Experience real-world messiness (latency, sensor noise, actuator limits)
- Validate sim-to-real transfer (simulation matches reality?)
- Build production systems
What you cannot do:
- Iterate as quickly (real robots take time to test)
- Recover from software failures as easily (crashing code in simulation is free; in reality it may damage hardware)
Timeline for Module 1: ✅ Full module (foundation, no motor control yet) Timeline for Module 2: ⚠️ Partial (can simulate in Gazebo, some tests on real robot) Timeline for Module 3: ✅ Full module Timeline for Module 4: ✅ Full module, with real-world validation
Cost: $1,500-$150,000+ depending on robot choice
Hardware Tier Comparison Table
| Aspect | Tier 1 | Tier 2 | Tier 3 | Tier 4 |
|---|---|---|---|---|
| Equipment Cost | $0-50 | $300-1,500 | $200-3,000 | $1,500-150,000+ |
| ROS 2 Local | Cloud | ✅ | ✅ | ✅ |
| Gazebo Simulation | Cloud | ✅ (local) | Partial | ✅ + Real |
| Iteration Speed | Slow (cloud) | Fast (local) | Medium | Slow (real hw) |
| Module 1 Possible | ✅ Full | ✅ Full | ✅ Full | ✅ Full |
| Module 2 Possible | ⚠️ Partial | ✅ Full | ✅ Full | ✅ Full |
| Module 3 Possible | ⚠️ Limited | ✅ Full | ✅ Full | ✅ Full |
| Module 4 Possible | ❌ (hard) | ⚠️ Difficult | ✅ Full | ✅ Full + Real |
Tier Progression Paths
Starting Tier 1 (Cloud Only)
Tier 1 (Weeks 1-5)
↓
Decision point:
- Continue Tier 1 (cloud path) for Modules 2-4
- Upgrade to Tier 2 (buy GPU laptop) for faster iteration
Recommendation: Complete Module 1 in Tier 1, then decide if cloud iteration speed is acceptable for Module 2. If frustrated by slow feedback loops, upgrade to Tier 2.
Starting Tier 2 (GPU Workstation)
Tier 2 (Weeks 1-8, Modules 1-3)
↓
Decision point:
- Continue Tier 2 for Module 4 (simulation-only humanoid)
- Upgrade to Tier 3 (buy Jetson) to test real robot constraints
Recommendation: Tier 2 is sufficient for learning Modules 1-3. Upgrade to Tier 3 if you want to test edge deployment before Module 4.
Starting Tier 3 (Jetson Edge Device)
Tier 3 (All modules possible)
↓
Optional: Add Tier 4 (physical robot) for real-world validation
Recommendation: Tier 3 covers everything. If you later want to test on real humanoid, add Tier 4.
Starting Tier 4 (Physical Robot)
Tier 4 (All modules, with real-world validation)
↓
Use Tier 2 or 3 for faster simulation iteration
Recommendation: Real robots are expensive and break if you make mistakes. Always develop in simulation first (Tier 2 or 3), then test on real hardware (Tier 4).
Cost-Benefit Analysis
Should You Upgrade to Tier 2?
Upgrade if:
- ✅ You find cloud iteration speed frustrating (10+ second wait per code change)
- ✅ You want to work offline
- ✅ You plan to spend 100+ hours on this course
Don't upgrade if:
- ❌ You only have a few hours to learn
- ❌ You're just exploring robotics (not committing)
- ❌ Your budget is very tight
Should You Upgrade to Tier 3?
Upgrade if:
- ✅ You want to deploy real edge code
- ✅ You plan to build actual robots
- ✅ You want to understand hardware constraints
Don't upgrade if:
- ❌ Simulation learning is sufficient for you
- ❌ You don't need real sensors
- ❌ Your budget doesn't support $1,500+
Should You Upgrade to Tier 4?
Upgrade if:
- ✅ You want to test on real hardware
- ✅ You have a specific robot project in mind
- ✅ You're in a research/professional context
Don't upgrade if:
- ❌ You're learning (simulation is sufficient)
- ❌ You can't risk hardware damage from software bugs
- ❌ Budget is constrained (robots are expensive)
Finding Your Tier
Quick Self-Assessment
Question 1: What equipment do you have access to right now?
- Only laptop without GPU → Tier 1
- Laptop with NVIDIA RTX GPU → Tier 2
- Jetson board → Tier 3
- Physical robot (Unitree, Boston Dynamics, etc.) → Tier 4
Question 2: How much time can you invest?
- Less than 10 hours → Tier 1 cloud is fine
- 50-100 hours → Consider Tier 2 for better iteration
- 200+ hours → Invest in Tier 3 or 4
Question 3: What's your learning goal?
- Understand robotics concepts → Any tier works (Tier 1 sufficient)
- Build working code → Tier 2+ recommended
- Deploy to edge → Tier 3+ required
- Real-world robots → Tier 4 eventually
What This Course Provides for Each Tier
Tier 1 Path
- Cloud ROS 2 setup guide (TheConstruct)
- Browser-based visualizations (MockROS)
- All code examples work in cloud terminal
- Slower iteration (5-10 seconds per test)
Tier 2 Path
- Local ROS 2 installation guide
- Gazebo/Isaac Sim setup
- All code examples work locally
- Fast iteration (1-2 seconds per test)
Tier 3 Path
- Jetson setup guide
- ROS 2 ARM-specific considerations
- Sensor integration examples
- Edge deployment patterns
Tier 4 Path
- Real robot setup guides (Unitree, etc.)
- Sim-to-real transfer validation
- Hardware safety patterns
- Real-world testing examples
Interactive Hardware Tier Selector
Below is a conceptual hardware profile assessment. In the real platform, this would be interactive (you select equipment, the system recommends paths).
Your Profile:
- Laptop specs: [CPU, RAM, GPU]
- Budget for upgrade: [$0, $500, $1,000, $5,000+]
- Time available: [Under 10 hrs, 50-100 hrs, 200+ hrs]
- Learning goal: [Understand concepts, Build code, Deploy edge, Real robots]
System Recommendation:
- Start with: [Tier 1/2/3/4]
- Can complete: [Which modules fully vs partially]
- Timeline estimate: [Weeks for each module]
Reflect
Think about your own situation:
- Your tier: Based on what you have right now, which tier are you starting in? Is that tier sufficient for what you want to learn?
- Upgrade pathway: If you wanted to upgrade tiers during the course, what's the minimal investment needed?
- Constraints: What's your biggest hardware constraint—budget, time, or something else? How does that affect your learning path?
These reflections help you make intentional decisions about hardware investment. Remember: simulation learning is valid. Many experts spend 90% of their time in simulation (Tier 1-2) and only 10% on real hardware (Tier 4). You don't need expensive equipment to learn effectively.