[Korea Aerospace University] Building an AI Development Environment for Autonomous Drone Research

For autonomous drone research in aerospace and mechanical engineering, Korea Aerospace University adopted an integrated AI development platform to provide an AI development environment and GPU resource management that many researchers can use at the same time.

A Multi-user AI Development Environment for Autonomous Drone Research

To conduct autonomous drone research including optimized MATLAB-based path computation, Korea Aerospace University needed an environment where many users could access the system at once and conduct research independently. It also required an operating foundation to save and manage each user's development environment and check GPU usage in real time.

Ensuring Research Continuity and Visibility into GPU Resources

The key challenge was to let many researchers develop reliably at the same time while preserving each user's research environment and efficiently managing GPU usage.

  • Multi-user development environment: An environment on DGX Station where many users can access it at once and develop independently

  • Saving development environments: An environment to save each user’s images and datasets so research can continue

  • GPU monitoring: A feature to check GPU usage in real time

Building an Integrated AI Development Environment on DGX Station

We deployed Cheetah on DGX Station to provide a development environment that many users can use simultaneously, together with dataset and volume management and resource monitoring.

  • Multi-tenant environment: Independent development environments that many users can access simultaneously

  • Dataset and volume management: Features to save and reuse development environments and data

  • Real-time resource monitoring: An intuitive UI to check GPU usage in real time

A Sustainable Research Environment and Efficient Resource Operations

Korea Aerospace University now has an environment where many users can develop at once, and researchers can continue their work while keeping their own MATLAB-based development environments and datasets. Intuitive GPU usage monitoring also improved the efficiency and visibility of research environment operations.

  • Concurrent research environment: Many users can access the platform at once and develop independently

  • Research continuity: Ongoing research by saving and reusing development environments and datasets

  • Greater operational visibility: Intuitive monitoring of GPU usage

Have more questions?

Contact us anytime to learn more about AIFRICA solutions.