GLA Summit

SOTA is evolving into a complete ecosystem for integrating and executing modern AI models directly inside LabVIEW.

After initially focusing on ONNX and ONNX Runtime, SOTA now supports a broader range of AI workflows, allowing LabVIEW developers to work with ONNX, PyTorch and SafeTensors-based models while keeping LabVIEW at the center of the application.

The objective is simple: make AI model integration in LabVIEW as straightforward as using any other software or hardware component.

With new drivers such as DeepMX, VisionMX and GraphMX, common tasks such as loading and executing an AI model, acquiring images from a camera, or running a computation graph can be configured quickly through simple LabVIEW interfaces and Express VIs.

DeepMX simplifies AI model execution.

VisionMX simplifies image acquisition and computer vision workflows.

GraphMX simplifies the execution of computational graphs and processing pipelines.

Behind these high-level interfaces, SOTA provides access to optimized runtimes and different execution providers, allowing applications to take advantage of the available hardware, from CPU execution to GPU acceleration.

This presentation will show how SOTA can be used to integrate modern AI into existing LabVIEW applications without requiring developers to redesign their software architecture around Python, external scripts or separate environments.

Through practical demonstrations, we will show how quickly a LabVIEW developer can:

- Acquire images from a camera

- Load and execute an AI model

- Work with ONNX, PyTorch and SafeTensors-based models

- Configure hardware-accelerated execution

- Integrate AI inference into an existing LabVIEW application

- Build complete acquisition, processing and AI workflows using SOTA

The goal of SOTA is not to create a separate AI environment around LabVIEW.

It is to make AI a natural component of LabVIEW dataflow programming.

With SOTA, integrating and executing modern AI models becomes a native part of the LabVIEW engineering workflow.