TL;DR

• AI is driving new computational demands at the industrial edge.
• Traditional GPUs compromise rugged industrial PC performance.
• MLSoC enables edge AI without sacrificing durability.
• Local LLMs reduce latency, cost, and cloud dependency.
• Edge AI enables greater autonomy, especially in remote sites.

Why this matters now

One of the most fascinating challenges facing manufacturers today is finding ways to bring emerging artificial intelligence (AI) technologies into the plant. It is becoming increasingly clear that AI software will redefine operations, helping deliver efficiency and value that will enable organizations to gain competitive advantage globally. However, that advantage comes with a significant requirement for computational power, creating new challenges for operational technology teams.

In a recent discussion between Emerson’s Gene Juknevicius and Manish Sharma and the editor of Robotics Business News, the pair explains that many facilities rely heavily on industrial PCs to run software at the industrial edge. These systems are designed to withstand harsh conditions such as vibration, shock, dirt, and extreme temperatures, making them essential to industrial operations.

However, modern AI software is pushing the limits of traditional industrial computing infrastructure:

“The traditional choice would be to bring in a graphics processing unit (GPU). But a GPU card dramatically reduces the ruggedization of the IPC. With a GPU, there’s a fan inside, temperature ranges are reduced, and the lifecycle of the equipment is shortened.”

This highlights the tradeoff between performance and durability in traditional approaches to edge AI.

Takeaway: Traditional GPU-based approaches challenge the ruggedness required for industrial edge computing.

To solve this problem, Emerson’s industrial PCs leverage Machine Learning System on Chip (MLSoC) technology to add AI-focused compute power at the edge without impacting ruggedization.

Takeaway: MLSoC enables AI acceleration at the edge while preserving industrial reliability.

Bringing LLMs local

One of the key advantages of physical AI enabled by MLSoC technology is the ability to bring large language models closer to the process. Today, most LLMs run in the cloud due to massive compute requirements, which works well for non-real-time applications like analytics and historical data processing. However, real-time use cases demand significantly lower latency.

“As soon as we can run these models locally, taking advantage of LLMs on the edge devices without needing to go to the cloud, it starts to become a better option for various applications. In effectively improving what we can do at the edge, we reduce cloud computing and dependency, while eliminating latency, security issues, and additional cost of sending data offsite for processing. That will create new opportunities for quality control, predictive analytics, autonomous facility surveillance and hazard detection, as well as opening doors to improvements on the factory floor we likely can’t even imagine right now.”

Running LLMs locally shifts AI from analysis to real-time operational impact.

Takeaway: Local LLM deployment enables faster, more secure, and cost-effective industrial AI applications.

A bridge to increased autonomy

Bringing LLMs closer to the edge has the potential to unlock significantly increased autonomous operations.

“Autonomous operation is typically being deployed first in remote locations where there are few or no people, as those are the locations where it typically has the greatest value.”

Remote environments introduce additional challenges, particularly around connectivity and data transmission.

“Typically, facilities will need some sort of wireless connectivity, and with machines moving in and out, the quality of wireless links can fluctuate quite a bit. Moreover, the amount of data users will need to send, especially as they inch closer to closing the loop, will be quite high.”

These constraints make cloud-dependent AI impractical for many autonomous scenarios.

Takeaway: Edge-based AI is critical for enabling autonomy in remote and connectivity-constrained environments.

These remote sites seeking more autonomous operation are ideal use cases for localized AI models running on rugged industrial PCs. Custom-trained models using technologies like vision systems can help improve quality, reduce scrap, and drive efficiency.

Takeaway: Rugged edge AI platforms enable practical, high-value autonomous industrial operations.