How should I choose the best vibration sensor for equipment maintenance?

Selection depends on equipment criticality and inspection frequency. While wired sensors are for ultra-critical assets, Wi-Fi wireless sensors that provide raw waveform data are optimal for automating inspections of general pumps and fans while enabling precision diagnostics.

What is the advantage of prioritizing raw vibration data?

The biggest advantage is that engineers can verify FFT waveforms to pinpoint root causes like bearing damage, rather than relying on black-boxed AI scores. It also serves as high-quality training data for building independent predictive maintenance systems using Python.

What anomalies can be identified from vibration waveforms (FFT)?

You can identify typical failure causes in rotating machinery by analyzing waveform patterns, such as unbalance (strong peak at rotation frequency), misalignment (integer multiple peaks), and specific high-frequency anomalies caused by bearing damage.

Practical Technical Guide for Vibration Monitoring

Practical Guide: Solving On-site Questions and Concerns

Vibration monitoring is a field where the gap between theory and actual site conditions can be vast. In this category, we go beyond basic concepts to provide a practical technical guide for engineers and maintenance professionals. From selecting the right sensor to analyzing complex waveforms, these articles offer the hands-on expertise required to achieve reliable, data-driven machinery diagnostics.

Engineering Insights: Selecting Sensors and Analyzing Waveforms

The first step toward successful monitoring is choosing a sensor that fits your specific environment, whether it be handy, wired, or wireless. To support this decision, we provide a clear selection guide and an extensive vibration waveform library that showcases actual signs of machinery anomalies. By comparing real-world waveforms of bearing damage and unbalance, engineers can learn how to identify potential failures with confidence. Our technical guides also delve into advanced data utilization, showing how to move from simple CSV exports to professional-grade Python analysis, empowering engineers to build their own predictive algorithms.

The Developer’s Perspective: Why We Prioritize Raw Data

At the heart of conanair’s development is a strict philosophy formed by field experts who understand the frustrations of "black-boxed" analysis. In our developer columns, we share the story of how conanair was built specifically to solve the limitations of conventional IoT sensors. We explain our technical choice to prioritize saving raw vibration data over simplified AI scores, ensuring that on-site engineers always have access to the facts behind an anomaly judgment. These insights reveal the "reverse thinking" and academic collaboration that make our low-cost, high-precision technology possible, providing a behind-the-scenes look at a sensor designed by maintenance experts for maintenance experts.

Building an "Open Maintenance System" for the AI Era

Rather than relying entirely on pre-packaged, closed-loop systems, we provide the foundational concepts for using your acquired raw data as high-quality training data for machine learning and AI model construction. We support your first steps in avoiding vendor lock-in and developing proprietary predictive maintenance algorithms tailored specifically to your equipment's unique characteristics.

A Data-Driven Approach to Successful Maintenance DX

Through the articles in this category, you will learn practical steps to resolve on-site challenges and build a predictive maintenance system that delivers real value. Use these insights to break free from black-boxed solutions and accumulate proprietary diagnostic knowledge within your own facility.

Select the ideal vibration sensor based on equipment criticality and site environment, balancing cost and performance without hesitation.
Understand real-world FFT waveform patterns for unbalance and bearing defects to enhance your team's primary diagnostic skills on the shop floor.
Leverage exported raw data (CSV) with Python and existing BI tools to build an independent AI predictive maintenance platform free from vendor lock-in.