Designing reliable electronics in just seconds

4 mins read

Circuit Mind, a UK start-up, is getting a lot of attention as it looks to use AI to automate the design of electronic circuits

According to co-founder and CEO Tomide Adesanmi Circuit Mind is using artificial intelligence (AI) to completely automate the design of electronic circuits and speed up the design of devices whether that’s robots, autonomous vehicles or advanced medical systems.

“Today, when it comes to creating electronic devices it takes teams of experienced engineers months, or even years, and involve what can only be described as painful, manual and tedious tasks,” says Adesanmi.

“Research suggests that every year over £40bn and 1.5bn hours are spent globally on circuit board design work. What we at Circuit Mind are looking to build is an extremely complex artificial intelligence that will be able to take into account all the requirements needed for a circuit board and then output a final board design which will be ready for manufacture.”

According to Adesanmi, this will not only speed up the design process but deliver better circuits at a fraction of the cost.

An electronic systems engineer with a Master’s from the University of Leeds, before starting Circuit Mind with co-founder Basilio Gentile, Adesanmi built high integrity AR displays for pilots at BAE Systems.

“I was involved in designing extremely complex systems but found that much of the software that we were using was dated, and it could takes months to complete a design project simply because you lacked the tools to build things faster,” he remembers.

He wanted to address this and took the decision to leave BAE Systems to create his own business to try and solve this particular problem.

“I wanted to give engineers the tools they needed to accelerate their work. Too often in companies the design process is well established and software vendors tend not to be incentivised to build better software. There is, quite understandably, a tendency among engineers that if it works why change, especially if that involves complex systems and there are high cost associated with switching.”

To address those concerns Circuit Mind is using AI to completely automate the design of electronic circuits.

“We want to help electronics engineers and design teams by building an intelligent platform that will enable them to speed up their work a hundred-fold and to develop electronic circuits from concept to prototype or production – our focus is on rapidity but also reliability.”

Circuit Mind looks to take a basic description of an idea; draft the requirements, select components, develop the circuit schematics and layouts and then connect design engineers with manufacturing and assembly partners to get those designs manufactured.

Co-Founder Basilio Gentile has two Master’s degrees and is an expert in mathematical algorithms for complex systems and has driven the development of the high-performance algorithms behind Circuit Mind’s solution – ACE.

High performance algorithms

“ACE helps electronics engineering teams by automatically generating, optimising, maintaining and redesigning component selection and electronic schematics in seconds,” explains Adesanmi,” and it does all this without errors. In future we want ACE to be able to actually generate PCB layouts.”

According to Adesanmi ACE enables users to specify requirements as a hardware architecture.

“The idea is that by providing the details you care or are concerned about, ACE will then figure everything else out. Our algorithms will add converters, choose components and add passives, for example, to create the perfect circuit for your needs.”

Crucially, ACE gives the engineer full control of the design process.

“We describe it as being a canvass for the engineer to work on,” explains Adesanmi. "“You are able to input all the time whether that’s adding a microcontroller or more memory, for example. We recommend that people create at least four iterations of their design in order to create a balanced design.”

ACE has been designed to enable the designer to generate multiple design options, and they can then trade off issues concerning size, cost, and power. ACE also uses the latest procurement information to ensure availability and indicate sources for components on the BOM.

“Then at a click of a button, ACE is able to generate a functional circuit from your high-level requirements, completing the entire design in seconds or minutes – rather than hours or days. And these designs are mathematically generated and error-free - the platform executes a series of independent rule checks on the circuit design to guarantee an errorless design,” adds Adesanmi.

The solution can be used by any engineer working across any sector and according to Adesanmi the reaction from engineers has been ‘amazing’.

“I think they’ve been surprised by what the platform can do for them and last month we started to deploy it with a number of customers. We’re currently talking to customers in consumer electronics and electronic manufacturing services but the range of sectors is enormous.”

As ACE learns from the inputs that are made in to it so, over time, it will be able to take on more repetitive or tedious tasks.

“For example, if you consider the different regulations and standards that need to be met, ACE will be able to deal with those too. The engineer will simply need to be a ‘decision-maker’ as the underlying data will be able to make all those decisions for you,” explains Adesanmi.

“Designing hardware is hard and with ACE we wanted to help engineers to more easily assess and balance all the requirements that are needed to produce the ideal design; we want to remove the delays and risk and to end the need for engineers to shuttle between distributor sites, datasheets, and ECAD tools. Engineers will no longer need to perform the tedious, but necessary tasks such as analysing power, noise margin, FMEAs and so forth.”

Circuit Mind has drawn considerable attention and this year it was chosen by Deeptech Labs (DTL) as one of five of the most promising deep technology companies in Europe, and DTL will fund the company as one of the inaugural cohort of its accelerator programme.

DTL is backed by pre-eminent names in the deep technology industry such as ARM, Cambridge Innovation Capital, Ewan Kirk, and the University of Cambridge and companies in the programme will receive £350,000 in funding from investors.

Another backer, Google for Startups, through its Black Founders Fund, is helping the business supporting technology development, as well as sales and marketing over the next year.

“The range of backers and the support we are receiving has been incredible. I’m very excited and we’ve barely started. There is so much more we can do and build,” enthuses Adesanmi.

“When we show people this platform they want to try it and give it a go. Companies want to get to get product to market faster and they want to improve quality and raise productivity and we are able to deliver on all of those.”

Adesanmi ends our discussion by talking about his vision for the company and it focuses on the idea of ‘democratising’ the design process.

“I would love to think that in 10 years’ time a student, using our platform, anywhere in the world will be able to jump on to a computer, describe an idea and see it realised in a matter of days where and when they want.”

A great ambition for the future of electronics design.