NEPI for Oregon UAS Accelerator Cohort 2026

Plug-and-play Edge AI software to help UAS teams rapidly prototype, test, and deploy flight-ready autonomous solutions

Why NEPI Is Included in the Accelerator

Rapid prototyping that helps teams get to a working solution faster

Instead of spending early months and dollars building foundational software to test and demo their concepts, Oregon UAS Accelerator teams can leverage NEPI software along with all of the tutorial and community recourses to jump start their efforts. This removes the need to create custom software infrastructure for hardware interfacing, data handling, edge AI orchestration, system automation, and user interfaces during the early stages of their efforts, allowing a single developer to accomplish in weeks what use to take an entire team of software engineers months if not years to accomplish.

Throughout the cohort, teams can use NEPI as an experimentation toolbox to get validate, test, and demonstrate their solution, then optimize their NEPI developed solution into a product ready system they can start delivering to customers.

For teams who want a deeper technical overview of NEPI’s architecture and smart data capabilities, you can explore the NEPI white papers below:

Smart Data Solutions White Paper
https://numurus.com/whitepapers-smart-data-solutions/

NEPI Software Platform White Paper
https://numurus.com/whitepapers-nepi-software/

Oregon UAS Accelerator cohort companies receive access to:

  • A Ready-to-Use NEPI software environment available via the NEPI Container

  • Access to NEPI Documentation and community support provided online.

  • Access to NEPI support and engineering services offered by Numurus to accelerate development.

  • Access to NEPI source-code repo for solution customization and deployment.

  • Access to NEPI Discord Channel setup for the Oregon UAS Accelerator community.

  • Option to become a NEPI GitHub contributor and help shape the software project.

  • 5 free Commercial NEPI Licenses after providing structured feedback on their experience using NEPI.

This offering is intended to give teams a practical, ready-to-use software platform while still allowing flexibility in how NEPI is applied to each teams project.

 

Teams commonly use NEPI to:

  • Run AI models at the edge
    Deploy computer vision and AI models onboard the aircraft for object detection, tracking, inspection, mapping, or situational awareness without relying on the cloud

  • Bring up live sensor and camera feeds
    Stream and visualize cameras, LiDAR, GPS, IMUs, and other sensors with overlays for debugging, validation, and mission awareness

  • Create low-code autonomy and automation workflows
    Trigger actions based on AI outputs or sensor conditions, such as detect, decide, and act pipelines, without needing a full autonomy software team

  • Integrate and manage multiple sensors quickly
    Use NEPI’s built-in drivers and tools to connect common UAS payloads and sensors without weeks of custom integration work

  • Run everything onboard an edge computer
    Operate fully in-field on platforms like NVIDIA Jetson, enabling real-time processing in GPS-denied or bandwidth-limited environments

You don’t need a large engineering team to build intelligent UAS systems.
NEPI helps startups move from prototype to field-ready systems faster.

How does NEPI help with rapid prototyping specifically?

NEPI removes the need to build foundational software for:

  • Plug-and-Play Hardware Integration System for cameras, navigation, pan-tilt, and other common solution components.

  • Data Management System for collecting, routing, and synchronization system wide data.

  • Onboard AI Management System for wrapped development, deployment, and orchestration of custom AI models.

  • Low-Code Automation and Event System for rapid application layer development and testing.

  • Device Configuration System for management of network, time, software, drivers, applications, and other system level functions.

  • Complete ROS-Base API interfaces for all system data and control components for autonomy development

This lets teams focus on testing ideas in flight, adjusting behavior between runs, and validating use cases faster.

What role does NEPI play in your system architecture?

NEPI acts as the integration and orchestration layer between sensors, onboard computing, AI models, and mission logic. It sits alongside your existing flight stack and payload software, handling data flow, automation, and edge AI execution without forcing you to redesign your system.

Do you have to commit to NEPI long term?

No. NEPI is provided as a tool to accelerate all parts of a teams solution development with no lock-in requirement. Teams can use it where it adds value and disengage or extend it after the program based on their own roadmap.

Can we run edge AI onboard, or is this just data collection?

NEPI is designed for onboard edge AI. Teams can run AI models directly on the vehicle during flight, trigger actions based on results, and evaluate outcomes immediately rather than relying only on post-processing.

Does NEPI replace our flight controller or autonomy stack?

No. NEPI does not replace flight control, navigation, or autonomy frameworks. It complements them by handling providing a simple backseat driver interface to NEPI’s  AI and automation processes.

How flexible is NEPI as our system evolves?

NEPI is built for iteration. Teams commonly adjust sensors, AI models, automation logic, and data workflows by tweaking the existing NEPI source-code, or adding new solutions to the base NEPI solution.

What hardware and compute platforms does NEPI support?

NEPI runs on common onboard compute platforms, including NVIDIA-based systems, and supports a wide range of sensors, pan-tilt, navigation, and other common solution hardware components. The platform is hardware-agnostic and designed to adapt to different airframes and payloads through a library of API interface classes, allowing different subsystem development efforts with minimal coordination between teams. 

Is NEPI only useful for computer vision use cases?

No. While many teams use NEPI for vision-based workflows, it also supports:

  • Multisensor fusion

  • Telemetry-driven automation

  • Event-based data capture

  • Mission logic tied to onboard analytics

How steep is the learning curve for teams new to edge AI?

NEPI is designed so teams don’t need to be edge-AI experts to get started. Much of the integration and orchestration is already handled, allowing teams to focus on applying models and testing behavior rather than building infrastructure.

What kind of data does NEPI help us collect?

NEPI supports structured, time-synchronized, nav-synchronized data collection and processing across sensors, AI outputs, and system events. This makes it easier to evaluate performance, compare test runs, and support reporting during and after the cohort.

Where can we learn more or go deeper technically?

Teams can explore documentation, tutorials, and examples on the NEPI website:
👉 https://nepi.com

Get Started With NEPI

NEPI Step-by-step process:

  1.  Visit the NEPI Container page and follow the step-by-step instructions to install and run the NEPI container.

  2.  Create an account on the NEPI Community page to access documentation, ask questions, and share feedback with the NEPI team.

  3.  Join the NEPI Discord channel for quick questions and discussion with other companies in your cohort.

  4. Use NEPI during the cohort as a common software foundation throughout the accelerator to support sensor integration, onboard AI, and mission automation as your system evolves.

  5.  Complete the short feedback survey to receive up to 5 free commercial licenses.

                         *This program is available exclusively to the registered 2026 Oregon UAS Accelerator Cohort *