Vibiscus AI-Powered Adaptive Membrane HVAC Noise Absorber

Design Solution · Acoustic

Design Solution · Dream about it

AI-driven active membrane system targeting low-frequency HVAC noise via real-time duct repositioning.

Vibiscus deploys networked in-duct microphones and machine learning to dynamically adjust physical membranes, addressing the 'acoustic gap' where passive absorption and conventional active noise cancellation fail: frequencies below 1 kHz in occupied buildings. Intended to reduce HVAC rumble and meet noise regulation in commercial and residential settings without requiring ductwork redesign.

Vibiscus targets a real performance gap: passive absorbers and conventional active noise cancellation both struggle with the low-frequency HVAC rumble below 1 kHz that tenants actually complain about, and retrofitting ductwork to eliminate it mechanically is expensive. The concept — in-duct microphones feeding machine learning that dynamically repositions physical membranes — is mechanistically coherent and addresses a genuine occupant-comfort and regulatory compliance problem in commercial and residential buildings. The claimed 40 dB reduction would be transformative if validated, but no third-party acoustic testing, no reference to an ISO or ASTM test standard, and no named customer installation is documented anywhere on this record. The adaptive membrane approach also introduces real maintenance considerations: moving parts in a duct handling variable load, temperature swings, dust, and long-term sensor drift are a different maintenance profile from passive systems. Algorithm behaviour under sensor failure or feedback instability has not been demonstrated. This sits firmly in the early-commercialisation category — a compelling proposition with a plausible mechanism, waiting for independent field data before a specifier can treat the performance claim as anything other than a starting hypothesis.

Strengths

• Targets low-frequency HVAC noise (below 1 kHz), a known weak point for passive and conventional ANC solutions
• Retrofits into existing ductwork without major mechanical redesign
• Adaptive real-time response to variable noise conditions rather than fixed tuning
• Addresses occupant comfort and regulatory compliance (noise limits) in dense urban/commercial contexts

Considerations

• Introduces moving mechanical parts (membranes) into ducts, increasing maintenance complexity and failure points (moderate)
• Requires continuous power and computational infrastructure (microphones, controllers, algorithms) for operation (moderate)
• Not suitable for systems with high dust/contamination or where duct access for sensor placement is limited (low)

Risks

• No published independent field trial data, third-party acoustic testing, or named customer installations; 40 dB reduction claim lacks ISO/ASTM test standard reference or installation context (critical)
• Unvalidated performance in real-world HVAC duty cycles (variable load, temperature swings, long-term sensor drift) (high)
• AI algorithm behavior under edge cases (sensor failure, feedback instability, interaction with existing VAV/damper systems) not yet demonstrated (high)
• Membrane actuation speed and power consumption under sustained low-frequency excitation not yet characterized (moderate)

Performance

• Claimed noise reduction: Up to 40dB (unverified, no standard cited)
• Target frequency range: Below 1kHz
• Seed funding raised: $2.19M (July 2023)
• Research origin: 8 years at FEMTO-ST (CNRS)

Reality check

Company founded 2021, spun from 8 years of CNRS/Université de Franche-Comté research. Raised $2.19M seed (July 2023) from UI Investissement and Deeptech Founders. Website updated Feb 2026. No peer-reviewed publications, case studies, pilot results, or customer names found. The 40 dB reduction claim appears in marketing without ISO 3744/3745 or equivalent field measurement standard. No evidence of commercial deployment, retrofit compatibility testing, or long-term reliability data in HVAC service environments.

#active_noise_cancellation #HVAC #low_frequency #AI #occupant_comfort #adaptive_control

Source