Your browser is no longer supported. Please upgrade your browser to improve your experience.

Looking for a software estimate? Let’s chat!

Bluefruit Software has been awarded £100,000 in AeroSpace Cornwall funding to carry out research and development into edge AI in embedded systems.

What’s edge AI? As we explained in our blog post about how to bring edge AI to embedded systems:

Edge AI is about AI-enabled devices and systems making the most of the information available to them to help make intelligent decisions or to provide critical feedback to human operators; and doing all of this without having to send that computation elsewhere for it to happen.

Why is edge AI such a big deal?

Until recently, AI systems working on embedded devices needed to send input data processing and analysis to data centres. Due to this, AI-enabled embedded software and hardware hasn’t benefitted from true real-time computing, with delays to data transfer potentially causing operational and safety issues. After all, embedded systems are operating at the mercy of physics, responding to conditions in real-time.

And not all embedded systems can have the luxury of sending data to be analysed, because they’re operating in low connectivity environments, or security is an issue.

Edge AI brings the benefits of AI-enabled analysis and decision making and puts it on systems that can’t afford delays in responses or have low to zero connectivity to broader networks.

Importantly, it also means safer devices for a wide range of sectors, including manufacturing, aerospace, medtech and more.

Edge AI R&D at Bluefruit Software

Our R&D work started with our Audio Classification Equipment (ACE) project. ACE found us creating a proof of concept that showed it was possible to categorise sound picked-up by an embedded system and classify it on the device itself. ACE, as an embedded software application, was built on a machine learning (ML) algorithm trained on small data sample sets to then classify diagnosis sounds into pre-set categories on the ACE device. It categorises sound in real-time.

The potential applications for having real-time audio classification are immense. With an edge AI system of this nature working on an embedded system, it would be possible (for instance) to diagnose internal machine faults through sound recognition, in real-time. Having such a quick and remote diagnosis would enable internal machinery faults to be detected earlier and more safely.

Our engineers were grateful for the opportunity to use the supercomputing facilities at Goonhilly Earth Station’s AI Institute, which enabled our team to train ACE’s AI quickly and effectively. Training AI, regardless of whether it will live in the cloud or on edge AI-enabled system, takes a lot of computational power and resources. Being able to use the supercomputing facilities at Goonhilly meant that iterations of the AI could happen at a considerable pace.

It took eight weeks for our team to bring ACE to a standalone system, and as we look to investigate further ways to put edge AI on embedded devices, we’re hoping to continue working with Goonhilly.

Next steps for Bluefruit Software

ACE has come a long way throughout 2020. We’re currently fielding inquiries from companies working in aerospace, biopharma, and energy industries to see how edge AI could be a part of their devices.

Caitlin Gould, Director at Bluefruit Software best sums up how we see the potential for edge AI:

“We believe using AI in this way, built into equipment, can make the entire device or machine more intelligent, more effective, and possibly safer for end users. For the companies we talk to, these are business-changing conversations. They would have the ability to have products that could predict faults before they become a catastrophic failure. This creates not only a better machine; it saves the company huge amounts in maintenance and diagnostics costs. We genuinely believe the application of edge AI in places where AI isn’t at the minute—places without the internet even—will enable it to change our daily lives; that is why we are so keen to explore and develop this technology.”

If you’d like to learn more about our work in edge AI, why not take a look at some of our links below?

How can we help?

For the past 20 years, Bluefruit Software has worked closely with its clients to deliver high-quality embedded software and firmware solutions across a range of sectors. Whether you’re looking to bring edge AI to a product or looking to update a legacy system, our team can help.

Set up a call with our engineers