Starkey gives an edge: Key features of its ‘best in industry’ Edge AI hearing aids

Starkey’s new Edge AI hearing aids have hit Australian and New Zealand shores with a webinar launch in February 2025 and regional roadshows around both countries. Managing director of Starkey Hearing Technologies ANZ, audiologist DAWN ROLLINGS, reveals key features.

Speech clarity including hearing in background noise is the number one benefit people seek from their hearing aids. Starkey’s newest technology, Edge AI, well and truly delivers on this, boasting 30% better speech identification than existing hearing aids.

‘Incredible feedback’ from patients wearing Edge AI overseas highlighting ‘amazing sound quality and connectivity’ led the company to bring forward its Australian and New Zealand launch by several months, says Starkey Hearing Technologies ANZ managing director Ms Dawn Rollings.

Starkey hastened getting the technology, which it says is the best yet, to patients here so they could experience its ‘pure, clean sound’ and seamless processing as soon as possible, she adds.

“Edge AI supersedes Genesis AI with better sound quality, improved signal-to-noise ratios, and a two-fold increase in streaming range,” Rollings says. “DNN processing is 100 times more powerful, enhancing sound clarity and streaming range.

“There’s also improved signal-to-noise ratios in noisy environments with a 6 dB reduction in low-level noise while maintaining the industry’s best 51 hours’ battery life.”

The device features the industry’s first DNN-powered speech probability predictor which leads to the 30% more accurate speech identification. Deep neural networks or DNN is a machine learning (ML) algorithm that uses multiple layers of neural networks to solve complex problems.

“We’re most excited about further enhancements to the sound quality and the way we achieve superior speech ID and sound quality using deep neural networks via the Neuro Processor,” Rollings says.

“The driving force of Edge AI is better sound quality through this unique integrated neural processing unit (NPU) which incorporates artificial intelligence (AI) via DNN technology.

“This is what people want. They want to be able to hear clearer in background noise without having to make multiple adjustments. The DNN uniquely can filter that for them.”

If someone is in a complex listening environment such as a restaurant or somewhere where there’s a lot of other speech and background noise, that’s where the significant incremental difference is, because the speech probability predictor is 30% more accurate at identifying speech across even more complex listening environments, she adds.

“The predictor basically classifies a lot of what we would call noise or speech babble, and makes it a very clean signal, so the listener can really hear the difference,” Rollings adds.

“It’s trained on millions of sounds to distinguish those of interest, such as someone talking to you, from distracting noises such as clanks, hums or other people’s conversations. The AI focuses on what we want to hear versus what we don’t want to hear, no matter where you are.”

Mimics the brain

The DNN technology mimics the way a brain normally processes information to not only improve sound clarity but also reduce listening effort and fatigue, Rollings says.

Starkey explains it as listening being a complex process that a normal-hearing brain handles effortlessly, which breaks down with hearing loss. “Edge AI’s technology mimics the brain’s auditory cortex to help fix that broken process, classifying complex soundscapes, enhancing speech and reducing noise in real time with advanced and powerful AI at the edge,” it says.

Another unique feature is that Edge AI has one DNN chip in one neuroprocessing unit rather than in multiple units which competitors’ hearing aids have, Rollings claims. It’s not a standalone unit sitting on the chip but is fully integrated into it, she adds.

“We’re the only ones who have the DNN chip integrated into one neural processing unit (NPU) which means we can increase features.

“By using this powerhorse we can mimic the brain and classify speech and speech babble – what we want to hear versus what we don’t want to hear,” she says.

This all happens automatically so users don’t need to make adjustments, Rollings says.

However, they can be made on demand in the MyStarkey app if needed and if someone wants to change sound optimisation or their preferences they can.

Apart from additional features, the neural processing unit’s neuroscience-inspired technology and powerful DNN also allow for better power management.

“It does all this while maintaining the industry’s best 51 hours’ battery life in a single charge for our RIC (receiver in canal) RT,” Rollings says.

Seamless connectivity

AI brings the device into the automatic realm and along with the DNN, everything is seamless including connectivity.

There’s enhanced connectivity with LE (Low Energy) Audio and Bluetooth plus Auracast connectivity and a new smaller TV Streamer for superior streaming.

Bluetooth and LE audio enable seamless connectivity with more devices, less dropouts and a more consistent signal, Rollings adds. They pair more easily, connect more quickly and stream more robustly.

The two-fold improvement in streaming range because of LE audio also means that when streaming, there’s a much better range so the hearing aids are more efficient and use less power.

‘Sound quality is amazing’

“It’s been a year since we launched Genesis AI, and now we already have on the market our latest and greatest Edge AI,” she says.

“The reason we’re doing this is because we’ve seen the impact it had when it was launched in the US and across Europe.

“The feedback from patients in terms of sound enhancement and sound quality is so significant, we thought, let’s get it to market here sooner.

“Feedback has been incredible. Patients have been saying the sound quality is amazing compared to anything else they’ve ever heard.

“Patients have even gone from Genesis AI to Edge AI and are noticing sound enhancement.

“The connectivity is something else people are calling out because it’s very easy to use, it’s an improvement on how devices connect, and streaming quality is excellent for sound, music and connection to audio and Auracast.”

Balance Assessment Tool

Health and wellbeing features are also enhanced by adding a balance assessment tool via on board sensors that provide self-guided balance assessments to identify if the client is at risk of falling.

“We already have a fall alert that alerts loved ones if the patient falls, but we’re going to the next level and saying let’s try and prevent that fall,” Rollings says.

Apart from having an on-demand control centre via the MyStarkey iPhone app, Edge AI also now integrates with the Apple Watch for additional health monitoring.

Rollings says the DNN works through sensory, subconscious and conscious means.

The sensory part takes into account acoustic, motion and listening intent information and maps this out using embedded DNN.

“The subconscious part works like our brain, effortlessly processing in the background without needing energy from the listener which is why they don’t fatigue as much,” she says.

“The conscious part means that needs can vary for two patients with the same hearing loss.

“Algorithms built for the listening experience are complex and there is a lot of flexibility depending on the wearer and where they are.

“Wearers can use on demand Edge Mode + for increased comfort and clarity through the MyStarkey App.”

Best in industry

Edge AI hearing aids are modern, discreet, comfortable and can be used in all listening situations.

“I think people are always going to go for what’s best in industry which is what Edge AI is.

“It’s a solution for everyone, right through custom products, the RIC 312 and all the RIC styles,” Rollings says.

“It has more accurate speech identification, increasing signal to noise improvements in noisy environments, the best in industry battery life and our rechargables are waterproof to one metre.

“It mimics what we do naturally and when you listen to Genesis AI versus Edge AI, Edge AI sounds better, even better than we thought was possible.”

Webinars and recordings

People can view recordings of the endorsed Edge AI product launch webinar as part of a three-part endorsed webinar series focusing on Neuro Sound 2.0, future forward connectivity and an examination of the balance assessment tool, at starkeylearninghub.com.au.