Xylo Labs
Category: AI in Manufacturing
South Korean AI startup developing on-device predictive maintenance solutions that fuse sound, vibration, and operational data to detect industrial equipment failures before they occur Xylo Labs was founded in 2025. The company is led by Gwangseok An (안광석). Based in Busan, South Korea. Team size: 1-10. Total funding raised: $133,000. Latest round: Seed. Key investors include MYSC (엠와이소셜컴퍼니).
- Founded
- 2025
- Headquarters
- Busan, South Korea
- Team size
- 1-10
- Total funding
- $133,000
Value proposition
Xylo Labs prevents industrial equipment failures by analyzing acoustic, vibration, and operational state data together through on-device AI, reducing false positives that plague conventional single-signal approaches and eliminating cloud dependency for security and cost benefits
Products and solutions
XyloZero (facility monitoring platform for real-time acoustic/vibration/environmental signal analysis with edge inference), Motor Sound Diagnostics, Terminal Operations Console, Robot Learning & Motion Planning Solutions
Unique value
Combines acoustic, vibration, AND operational state data (speed, load, on/off status) to distinguish normal operational variation from actual anomalies — solving the false-positive problem that plagues conventional single-signal predictive maintenance; on-device AI eliminates cloud transmission costs and security risks; single acoustic sensor can monitor multiple nearby machines
Target customer
Port operators, power plants, manufacturing facilities, heavy industry, and shipping companies with large-scale rotating equipment requiring continuous condition monitoring
Industries served
Ports & logistics, power generation, manufacturing, heavy industry, shipping, forestry (pine wilt detection), robotics
Technology advantage
On-device AI (edge inference) for real-time processing without cloud dependency; multi-modal signal fusion (acoustic + vibration + operational telemetry); acoustic-sensor architecture enabling one device to monitor multiple machines; expertise in acoustic and signal AI from 6+ years of Seoul National University research; hardware-converged systems from sensor to operator software
How they differentiate
Unlike competitors that focus on single-signal analysis (vibration-only or ultrasound-only), Xylo Labs fuses acoustic, vibration, AND operational state data to eliminate false positives. Their on-device AI architecture avoids cloud dependency, and their acoustic sensor design allows one device to cover multiple machines, reducing installation costs. The company also has a unique background in forestry diagnostics (pine wilt detection) and voice security applications.
Main competitors
Movic Lab (모빅랩) - ultrasonic/acoustic signal analysis for industrial equipment, Futuremain (퓨처메인) - factory predictive maintenance using vibration data, AIBIZ (에이아이비즈) - AI-based anomaly diagnosis for semiconductor manufacturing
Key partnerships
Busan Port Authority (pilot for crane drivetrain monitoring), Korea Southern Power (pilot for coal transport facility diagnostics), Korea Midland Power (pilot for power station equipment monitoring), Seoul National University (research collaboration and founding base), Korea Forest Research Institute (field collaboration in tree diagnostics), MYSC (investor and accelerator via EMA-MARINE program)
Notable customers
Busan Port Authority (Busan New Port crane monitoring pilot), Korea Southern Power (coal transport facility diagnostics pilot), Korea Midland Power (power station equipment monitoring pilot)
Major milestones
2019: Research began at Seoul National University (SNU Field Problem Research Group), 2021: LG CNS Startup Monster selection, 2022: Pre-Startup Package selection by KISED, 2025-12: Company incorporated, 2026-06: Seed round of ₩2억 ($133K) from MYSC, 2026: Selected for MYSC's EMA-MARINE global track for Singapore expansion, 2026: Selected for Hyundai Motor Chung Mong-koo Foundation's CMK Impact Pruner program
Growth metrics
3 employees (as of June 2026 per THE VC); incorporated December 2025; active pilots with Busan Port Authority, Korea Southern Power, Korea Midland Power
Market positioning
Early-stage Korean deep-tech startup positioned in the industrial predictive maintenance market, targeting ports, power plants, and manufacturing. Competes against more established Korean players (Movic Lab, Futuremain, AIBIZ) with a differentiated multi-modal signal fusion approach and on-device AI architecture.
Geographic focus
South Korea (domestic focus with pilots at major port and power operators), with plans to expand via MYSC's EMA-MARINE global track (Singapore market entry)
Patents and IP
Signal-processing patents from Seoul National University research (2019-2025); specific patent details not publicly listed on website
About Gwangseok An (안광석)
PhD in Acoustic Signal Processing from Seoul National University Graduate School of Convergence Science & Technology; Former R&D Lab Director at a space launch vehicle startup overseeing hardware development; 10+ years of acoustic/signal processing research since 2019 at SNU
Latest news about Xylo Labs
More AI in Manufacturing companies
Official website: https://xylolabs.com
