)
:format(webp))
Phonak gathered about 300 hearing care professionals on November 11-12 to the NASA Johnson Space Center in Houston to introduce Infinio Ultra, a firmware-driven platform expansion that extends major new capabilities to every current Infinio hearing aid user. It was an ambitious setting for a product launch, but the symbolism wasn’t lost on the attendees.
“Today marks a turning point,” said Christine Jones, AuD, Sonova VP of Marketing Wholesale, US. “We have learned, we have listened… and everything we incorporated over the past year has brought us to the point of Infinio Ultra.”
:format(webp))
The Infinio Ultra platform’s new innovations include a more intelligent AutoSense 7.0, a completely retrained Spheric Speech Clarity 2.0, redesigned one-step connectivity, and a deceptively simple but important EasyGuard wax-protection system. Ultra also offers a giant leap forward in improving the battery life of the platform. During presentations and two exclusive interviews with HearingTracker, Phonak and Sonova leaders explained how Infinio Ultra works, its benefits for users and hearing care professionals, and why this latest firmware update may help reshape the hearing industry.
With more than 1.5 million devices sold in its first year, Infinio became Sonova’s most successful product launch. “I don’t know of any other product that has been asked by name as many times as Sphere,” said Jimena Garcia, Sonova Vice President of Commercial Sales. “Patients were calling clinics before they even had a hearing test, just to ask who had Sphere in stock.”
:format(webp))
The response didn’t just generate enthusiasm—it also yielded a treasure trove of real-world data for Phonak. With over a million Infinio and Sphere fittings worldwide, Phonak was able to retrain its Deep Neural Network (DNN) and automatic classifier with an unprecedented diversity of listening environments, speech samples, and user interactions.
This is why, as Jones emphasized during an interview, that AutoSense 7.0 is 24% more precise in how it classifies and responds to challenging acoustic environments: “We trained the system with 18-fold more datasets…AutoSense 7.0 is the most precise system we’ve ever had.”
AutoSense OS, described at the event as the “mission control” of Phonak hearing aids, sits atop a multilayered architecture. Brandy Pouliot, Director of Audiology and Education, uses a house analogy to explain it: “APD [Adaptive Phonak Digital] is the foundation. Above it are rooms—quiet listening, speech in noise, comfort in noise, car, speech-in-loud-noise, etc.—spaces where technologies like StereoZoom 2.0 and Sphere live and operate.” The “garage,” she noted, includes many other features like connectivity, streaming functions, as well as Roger accessory technologies.
What changes with Ultra is how confidently AutoSense moves between rooms. With substantially more training data, AutoSense 7.0 can detect subtle shifts earlier and apply the appropriate blend of gain models, noise reduction, directionality, or DNN processing with fewer misclassifications.
In a structural sense, the new APD 3.0 is a fundamental upgrade. Pouliot emphasized that in Phonak trials, APD 3.0 was preferred 93% of the time over competing premium sound-quality systems: “It’s our signature sound—third generation—and the foundation of everything. It’s why people say Phonak sounds like Phonak.”
While AutoSense governs when and how features activate, Spheric Speech Clarity 2.0 (SSC 2.0) determines how well speech is actually understood—especially in noisy and dynamic settings. “Beamformers always had a limitation for spatial perception,” Stefan Launer, PhD, Sonova VP of Audiology and Health Innovation, told HearingTracker. “With DNNs, you don’t have this limitation anymore. You can have access to speech sources from all around… and get exceptional SNR improvement.”
Unlike directional microphones, which must be pointed at a talker, SSC 2.0 directly extracts speech from background noise using a DNN with millions of parameters trained to mimic acoustic properties of real speech, including:
This is why users often leave Sphere turned on continuously — “far more than any previous data would’ve led us to believe,” Jones said.
Kevin Seitz-Paquette, AuD, Sonova’s Senior Director of Audiologic Insights, showed a video of a test participant surrounded by five speakers, each contributing one word to a sentence, with words coming from random directions in a noisy field. The results indicated a 44% improvement in correct word identification over StereoZoom and 37% reduction in listening effort (validated by behavioral ratings and EEGs).
“Every word was crisp and distinct,” Seitz-Paquette said. “Not because beamforming did a good job—but because Sphere directly extracted the speech.” In a separate study, participants described being more willing to attend social events, less reliant on spouses to “interpret,” and generally more confident navigating noisy environments.
Battery performance was one of the few areas where early Infinio users—and even some reviewers—hoped for more. Spheric Speech Clarity (SSC) is computationally demanding. Running DNN continuously in real-world environments can lead to high energy consumption. While Infinio delivered strong acoustic benefits, heavy SSC usage could shorten runtime to levels that some users found challenging. With Ultra, that changes dramatically.
Infinio Ultra’s power consumption has been reduced by 30%, directly translating into longer daily use—especially when being in SSC mode for extended periods. Launer was straightforward about the improvement: “With the first version, [continuous Spheric use] was 5-7 hours… Now, you can use it for 10-11 hours.” In fact, battery life exceeds 50 hours when not using Sphere or audio streaming, rivaling the efficiency of some of the most energy-miserly hearing aids on the market.
:format(webp))
He explained that Ultra’s improvement in battery efficiency comes down to something surprisingly simple: for the first time, the engineering team had the real chip in hand. After launch, Sonova engineers were able to directly measure the chip’s real convolution layers, recurrent layers, memory access cycles, and other components, allowing them to retrain the DNN around the actual electrical behavior.
Stefan LaunerWe could measure…all the functions, and we could optimize the DNN with the real hardware, not just the model. That let us scale the power consumption down by 30%.
Clinical testing further reinforced Phonak’s updated confidence. VP of Clinical Research Strategy Jason Galster says that software-based efficiency gains of this magnitude—delivered as a firmware upgrade—are essentially unheard of in the hearing aid industry, where major battery-life improvements almost always require a new chip, new battery chemistry, or a new housing. “Being able to provide somebody with a 30% improvement in power consumption simply through a firmware upgrade—I think that’s a first.”
In short, Ultra turns one of the Infinio platform’s only notable weaknesses into a defining strength, and it does so in a way that underscores the broader theme of this release: meaningful innovation delivered through software. This has the potential to reframe consumer expectations about product cycles and positions Phonak as one of the first major manufacturers to demonstrate that transformative improvements are increasingly occurring in firmware rather than hardware.
:format(webp))
Connectivity problems historically represented the top category of clinic troubleshooting visits. Ultra’s update aims to dismantle that. The Ultra update includes near-instant Bluetooth pairing, faster handshakes between multiple devices, and stronger reconnection logic.
“The ease with which you can connect patients and not have to do deep diving… it’s just so easy now,” Jones said. Garcia added: “Reducing friction and waste for hearing-care providers—that’s one of Ultra’s biggest contributions.”
The bottom line is that users can now move between home, car, office, café, and streaming environments with fewer interruptions and fewer manual adjustments.
Phonak’s EasyGuard system is a next-generation wax-management dome designed to reduce one of the most common sources of hearing aid failure: receiver blockage due to earwax. Although it may appear to be a simple silicone dome, EasyGuard represents a significant engineering upgrade over standard domes, combining enhanced acoustic transparency, improved wax resistance, and an optimized and more comfortable fit.
At its core, EasyGuard functions as a dual-purpose protective interface:
EasyGuard’s shape and flexibility come from thousands of 3D ear impressions to identify the most common canal shapes and pressure points. “We used AI and finite-element modeling to design the shape and geometry… and to make the material mechanically completely sealed to keep the wax out but acoustically transparent,” says Launer.
:format(webp))
In a study cited by Seitz-Paquette, participants compared EasyGuard to their existing domes and showed a strong preference for the new design—not just for its protection, but also because it was more comfortable in daily wear. “People clearly preferred wearing this new EasyGuard dome system over their existing domes,” he says.
Garcia called EasyGuard her favorite part of the Ultra release. “The biggest service appointments are related to wax,” she says. “This is going to be such a huge benefit for reducing waste in practices and their workflow.”
The event also highlighted the features of Phonak’s Virto R Infinio custom in-ear rechargeable line. As with the rest of the Infinio Ultra line, it includes motion-sensor hearing, speech-sensor technology for softer voices, and AutoSense 7.0. Sonova President of Hearing Instruments (North America) Nicholai Dessypris emphasized that Virto R delivers “audiological performance without compromise,” combining strong connectivity, the smallest rechargeable custom form factor, and a user experience that feels “equivalent to a RIC” while preserving the discreet appeal and comfort expected from a custom device. Garcia added that clinicians have found the product “super comfortable,” even for small ears, and that real-world streaming with Virto R is “better than the AirPods” in her experience.
:format(webp))
Phonak emphasized a new philosophy: the platform is the hardware; the innovation is the firmware. Jones put it plainly: “It’s unprecedented that someone who invested last year can automatically gain everything we’ve developed since.”
Ultra may represent a shift away from multi-year replacement cycles and toward continuous delivery of improvements like better battery life, updated DNN models, improved connectivity, and other changes (e.g., AutoSense retraining). For clinicians, this means fewer dissatisfied patients who bought the “latest and greatest” hearing aid only to find out a month later that a newer, better product has become available. For manufacturers, using real-world data from field-tested products can help them maximize the benefits of the costly chip development process. Finally, consumers who are early adopters of technology are more likely to benefit from future upgrades.
And, as Galster noted, firmware-driven progress appears to correlate with increased daily wear time, which may be the strongest indicator of acceptance: “One platform after another has incrementally higher wear time… an objective sign of success.”
AI isn’t new to Phonak. In fact, as Launer reminded attendees, the company shipped its first hearing aid with AI-based algorithms about 25 years ago in the Phonak Claro. But Infinio Ultra represents a qualitative shift—from AI supporting the hearing aid to AI driving the hearing aid.
:format(webp))
Phonak’s dual-processor architecture—a traditional DSP chip and a dedicated neural processing unit (NPU)— sets this direction. As Launer put it: “If you want powerful, large-scale DNNs that directly extract speech, you need a powerful separate chip… just like a GPU or graphics card in a gaming computer.” This design allows Phonak to:
The implication is that the competitive battleground may be shifting. The winners won’t necessarily be the companies with the newest chip every two years—but those who can make the most of mature AI training pipelines. “We’re translating meaningful outcomes for providers, patients, and clinics,” said Dessypris. “It’s not a bug fix… it’s transformational.”
So in this way, Phonak’s moonshot framing wasn’t hyperbole. Ultra pushes toward a vision where hearing aids extract speech like high-end audio processors, battery life improves through real-world experience and efficiency, and users get better performance every year—without necessarily buying new hardware.
As Launer said: “Houston, we have a solution.”
Karl Strom ist der Chefredakteur von HearingTracker. Er war Gründungsredakteur von The Hearing Review und berichtet seit über 30 Jahren über die Hörhilfenindustrie.