Setup Guide

Getting Started with ASPEN

From unboxing to first simulation

Executive Summary

What you need at a glance

⚙

Hardware

Dell Alienware 18 Area-51 with RTX 5090, 64 GB RAM, 2 TB SSD. Approximate cost: ~$4,500.

💻

Software

Ubuntu 22.04 LTS, NVIDIA Isaac Sim 5.1 (free), Python 3.10+ via conda. All dependencies are open-source or free-tier.

📅

Timeline

~2–3 weeks from hardware arrival to a working demo, assuming no procurement delays.

🎯

What You'll Have

A portable demo station running underwater vehicle simulations with a web-based C2 interface for mission planning and monitoring.

Section 1

Hardware Setup

Acquire and configure the demo workstation

1

Acquire the Alienware 18 Area-51

Order a Dell Alienware 18 Area-51 laptop with the following minimum specs:

  • GPU: NVIDIA RTX 5090 Laptop (required for Isaac Sim)
  • RAM: 64 GB DDR5
  • Storage: 2 TB NVMe SSD
💡 The RTX 5090 is the minimum recommended GPU. Isaac Sim is GPU-intensive and lower-tier cards will not provide usable frame rates with complex ocean environments.
2

Install Ubuntu 22.04 LTS

Install Ubuntu as a dedicated partition or dual-boot alongside Windows. A dedicated install is recommended for stability.

# Download Ubuntu 22.04.4 LTS from ubuntu.com # Create a bootable USB with Rufus or Balena Etcher # Boot from USB and follow installer prompts
⚠️ Ubuntu 22.04 specifically is required. Isaac Sim 5.1 has known compatibility issues with Ubuntu 24.04 and has not been validated on other distributions.
3

Install NVIDIA Drivers

Install the latest NVIDIA proprietary drivers. Version 560+ is recommended for RTX 5090 support.

sudo apt update && sudo apt upgrade -y sudo apt install -y nvidia-driver-560 sudo reboot
4

Verify GPU

After reboot, confirm the driver is installed and the GPU is detected.

nvidia-smi # Should display RTX 5090, driver version, CUDA version
💡 If nvidia-smi returns an error, try sudo ubuntu-drivers install for automatic driver selection.
Section 2

Software Installation

Python environment, Isaac Sim, and ASPEN SDK

5

Install Miniforge

Miniforge provides a lightweight conda package manager. This is the recommended way to manage Python environments for ASPEN.

curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-x86_64.sh" bash Miniforge3-Linux-x86_64.sh # Follow prompts, then restart your shell source ~/.bashrc
6

Create Conda Environment

Create an isolated Python 3.10 environment for ASPEN development.

conda create -n aspen python=3.10 -y conda activate aspen
7

Install NVIDIA Isaac Sim 5.1

Isaac Sim is the simulation engine that powers ASPEN's 3D ocean environment. It is free to use via NVIDIA's Omniverse platform.

# Option A: pip install (recommended) pip install isaacsim # Option B: Install from NVIDIA NGC container registry # See: https://catalog.ngc.nvidia.com/orgs/nvidia/containers/isaac-sim
ℹ️ The pip install is approximately 12 GB. Ensure you have sufficient disk space and a stable internet connection. First launch will download additional assets.
8

Install Git LFS

Git Large File Storage is required to pull the 3D model assets, environment textures, and other binary files tracked in the ASPEN repositories.

sudo apt install -y git-lfs git lfs install
9

Clone the ASPEN SDK

The main repository includes three Git submodules (core simulation, vehicle models, and environment assets) plus large binary assets tracked by Git LFS.

git clone --recurse-submodules https://git.codev.mitre.org/scm/aspenext/aspensdk.git cd aspensdk git lfs pull
⚠️ The --recurse-submodules flag is critical. Without it, the vehicle models and environment directories will be empty and simulations will fail to load.
10

Install Python Dependencies

Install all required Python packages from the project's requirements file.

cd aspensdk pip install -r requirements.txt
11

Install Bellhop Acoustic Toolbox (Optional)

Bellhop provides acoustic ray-tracing for sonar and communication modeling. This requires compiling Fortran source code.

# Install Fortran compiler sudo apt install -y gfortran make # Clone and build the Acoustics Toolbox git clone https://github.com/oalib-acoustics/Acoustics-Toolbox.git cd Acoustics-Toolbox make install
💡 Bellhop is only needed if you plan to run acoustic propagation simulations. The core vehicle simulations work without it.
Section 3

First Simulation

Run a pre-built example and verify everything works

12

Run a Pre-Built Example

Launch the Dabob Bay scenario with GPU-accelerated REMUS vehicles. This is the fastest way to confirm your installation is working.

conda activate aspen cd aspensdk python examples/Remus_Dabob_Bay_GPU.py
13

What You Should See

After a 30–60 second loading period, you should see:

  • An Isaac Sim window with a 3D ocean environment (Dabob Bay bathymetry)
  • One or more REMUS 100 vehicles spawned at their start positions
  • Vehicles navigating waypoints autonomously with visible thruster effects
  • A real-time viewport you can orbit, pan, and zoom with mouse controls
💡 First launch takes significantly longer as Isaac Sim compiles shaders and caches assets. Subsequent launches will be much faster.
14

Common Issues and Fixes

If the simulation does not launch correctly, check these common problems:

Symptom Likely Cause Fix
CUDA error on launch CUDA / driver version mismatch Run nvidia-smi and verify CUDA version matches Isaac Sim requirements. Update driver if needed.
Missing 3D models or textures Git LFS files not pulled Run git lfs pull in the repo root and all submodule directories.
Black screen / no display Headless GPU or display config Ensure you are running on a display-connected session (not SSH). Try export DISPLAY=:0.
Import errors Wrong conda environment Verify conda activate aspen and re-run pip install -r requirements.txt.
Extremely low FPS (<5) GPU not being used Check that nvidia-smi shows Isaac Sim using the GPU. Verify NVIDIA drivers, not nouveau.
Section 4

C2 UI Setup (Optional)

Web-based command-and-control interface for mission planning

15

Install Node.js 18+

The C2 UI is a React/TypeScript web application that requires Node.js.

# Install via NodeSource curl -fsSL https://deb.nodesource.com/setup_18.x | sudo -E bash - sudo apt install -y nodejs node --version # Should print v18.x.x or higher
16

Clone the C2 UI Repository

git clone https://git.codev.mitre.org/scm/aspenext/aspen-c2-ui.git cd aspen-c2-ui
17

Install Dependencies and Start Dev Server

npm install npm run dev
18

Open the C2 Interface In Development

Navigate your browser to the local development server.

# Open in your browser: http://localhost:5174
19

Note: Mock Data Mode

The C2 UI currently runs on mock data out of the box. It will display simulated vehicle telemetry, waypoints, and mission status without a live simulation connection.

ℹ️ The API bridge connecting the C2 UI to live Isaac Sim telemetry is under development. For now, the UI is useful for exploring the interface design and mission planning workflows.
Section 5

Next Steps

Where to go from here