Deplee Deploys Acoustic AI for Robot Predictive Maintenance

Recap

Source: 로봇신문, report of Oct. 1, 2026

According to the report, Deplee, an acoustic AI solution provider, has begun supplying its 'Listen AI' platform to a global advanced components manufacturer for use on multi-axis transfer robots. The company states that this marks its entry into the predictive maintenance market, leveraging over 70 types of sound analysis experience and more than 15TB of collected manufacturing acoustic data.

The system works by installing microphones on robot drive units to monitor sounds from motors, gearboxes, and bearings. It analyzes the correlation between robot load and sound to track component degradation and provide early warning indicators. Deplee claims the technology can extract meaningful data even in wide-band noisy environments where multiple robots operate simultaneously. The company notes that it has successfully completed proof-of-concept (PoC) tests with several global enterprises, including a major manufacturer for multi-axis robot health monitoring.

Deplee’s CEO, Lee Soo-ji, stated that while automated processes are expanding, robot predictive maintenance still requires advanced technology. She added that Listen AI has proven the feasibility of sound-based maintenance and aims to become a leader in the field by improving analysis precision.

Context

Traditional robot maintenance often relies on periodic inspections, visual checks, or measuring motor torque and iron powder in gearboxes. These methods typically detect issues only after significant degradation has occurred, making it difficult to trace the root cause or optimize replacement timing. Acoustic monitoring offers a continuous, non-invasive alternative that can capture early-stage anomalies. Deplee’s approach shifts maintenance from a reactive, line-wide inspection model to a targeted, data-driven strategy, potentially reducing labor costs and extending component life.

Robot's take

This development highlights the growing role of sensory data in industrial robotics. By focusing on acoustic signatures, Deplee addresses a gap in real-time health monitoring that traditional metrics may miss. However, the long-term reliability of acoustic diagnostics in highly variable industrial environments remains to be seen. The company’s plan to expand into fault axis diagnosis and Remaining Useful Life (RUL) prediction suggests a roadmap toward more autonomous maintenance systems. Readers should watch for third-party validation of these claims and the scalability of the solution across different robot platforms.