How EDF is using audio to monitor its nuclear power plant equipment

EDF Energy has turned to a small London-based startup to trial its technology as a way to safely monitor equipment and create a whole range of operational efficiencies within its nuclear power plants.

Chirp was founded in 2011 as a spinout from University College London (UCL), founded on a technology that it describes as an audio barcode: it encodes data into a series of audible or inaudible near-ultrasonic pitches and tones.

Once encoded, this data - which can be anything from an image file to an electronic payment - is transmitted over the air to a receiving device, where it is decoded.

The project

In January this year, Chirp and EDF Energy secured £100,000 worth of funding from Innovate UK to trial data-over-sound solutions in EDF's nuclear power plants, namely as a way of transmitting diagnostics data from its legacy industrial equipment.

See also: How an Uber-esque on-demand bus service uses 'data over sound' to verify passengers

EDF Energy first approached Chirp to run a proof of concept in November 2016 to take readings from a gauge in a nuclear power plant at a range of 50 metres over a twelve hour period.

After hitting a 100 percent data accuracy rate in that trial the two companies embarked on a new project in October 2017: testing this data transmission approach across a range of equipment in EDF's Heysham 1 power plant.

This starts with the two companies looking to gain sensor readings from a wider range of equipment and communicating them wirelessly.

Another use for the technology is transmitting status updates from workers to check their position around the facility in real time on a tablet. This is useful when operating a large workforce with a lengthy queue of jobs as managers get better visibility of the status of their staff, meaning the next worker can be queued up more quickly.

Chirp is also investigating transmitting data from wearables, and portable devices for taking readings from plant equipment.


There are a set of challenges EDF Energy faced before turning to Chirp, namely that its equipment predates radio frequency (RF), and the cost of implementing that technology for monitoring purposes would be prohibitively high. What the company essentially wanted Chirp to do was to retrofit that equipment to transmit data, without the cost associated with RF.

Then there's the acoustic challenges a cavernous nuclear power station provides, with a lot of background noise posing the risk of disrupting data transfers. It seems that Chirp's technology passed with flying colours during the initial proof of concept.

Previously, EDF Energy has been looking for a way to extend the lifetime of its 14 advanced gas-cooled reactors (AGR), with better access to data regarding its critical components outlined as a priority when it launched a funding competition with Innovate UK in July 2016.

These reactors operate in tricky conditions at temperatures of up to 650 degrees centigrade, making physical inspections tricky. EDF wanted technology that would allow more remote, but accurate inspection.

Despite being a separate initiative, there is some overlap with the work now being done with Chirp.

Dave Stanley, project manager in EDF Energy's innovation delivery team told Computerworld UK: "Where we do overlap is that with further development to shield the hardware from the effects of highly irradiated areas, Chirp gives the potential to take readings remotely from the same challenging conditions."

Read next: How EDF wants AI to optimise its nuclear power stations and the smart home

What next?

Chief science officer at Chirp, Dan Jones, explained to Computerworld UK that the company is building "an ad hoc sensor network with a data logger, transmitting data wirelessly via audio so that they can collect big data and information across the plant that would be too laborious or costly to connect [manually]."

"Chirp is agnostic to the data we communicate," he explained. "So we can take readings from anything that generates a reading, take an analog output and transmit that via Chirp instead of taking a physical reading."

The overall aim for EDF is to keep equipment running more of the time.

"Gathering more information from the status of equipment allows them to do more predictive maintenance and gather data for any further insights," Jones said.

For example, EDF plants are regularly put into an outage state for equipment to be maintained, so making those periods as short and effective as possible is a huge benefit to EDF.

"We see this as start of a longer and richer relationship with EDF," Jones said. "We want to explore these use cases and roll out across a much wider range of opportunities in the generation industry.

"The EDF support has opened our eyes to a lot more opportunities outside of nuclear where RF is prohibited. Take oil and gas, military and medical, so we see a lot of places Chirp could play a meaningful role."

Naturally this means Chirp will be expanding for the year ahead, with a new round of funding planned this year and, we presume, a whole lot of hiring.

Copyright © 2018 IDG Communications, Inc.

Bing’s AI chatbot came to work for me. I had to fire it.
Shop Tech Products at Amazon