Automation is changing the job roles of human employees. The days of rigorous and repetitive diagnostic checks are moving to the wayside to clear a path for a more autonomous future. Whether they work in car manufacturing, PCB fabrication, or even warehousing logistics, workers’ roles will become more about monitoring robotic systems than performing physical labor.
Clearpath Robotics Platform OTTO automates materials handling operations in a variety of warehouse environments. Image used courtesy of Clearpath Robotics
Condition-based monitoring (CbM) is an increasingly important strategy in Industry 4.0 and is becoming more widespread with the advent of open-source development platforms using Python and MATLAB.
Analog Devices recently announced a new hardware development kit designed for vibration-centric condition-based monitoring, the CN0549. The CN0549 kit interfaces with existing hardware processors, including FPGAs and microcontroller-based units compatible with the Arduino form-factor, thus allowing for design portability.
Why Condition-Based Monitoring is Critical to Modern Industry
Condition-based monitoring is a subset of the preventative maintenance paradigm to reduce costs and extend operational asset life. Planned maintenance is critical to maintaining equipment uptime and extending the life of the equipment. CbM increases efficiency and reduces cost by reducing unnecessary maintenance cycles.
Bearings, structural frameworks, and wheels can degrade under continuous load operations with mechanical and robotic systems. Vibrations can cause machines and devices to breakdown easier, which makes vibration-based diagnostic sensors, like the ADXL1002, an important part of designing durable, long-lasting robotic systems.
The basic paradigm of preventative maintenance and how planned maintenance and CbM interface with each other. Image used courtesy of Prometheus Group
Multiple active and passive techniques exist to facilitate CbM, including:
- Vibration characterization
- Oil analysis
- Sonic and ultrasonic detection
- Radiation and laser-based analysis
Vibration CbM and Analog Devices’ New CbM Development Kit
As a three-part kit, the CN0549 offers hardware designers, software engineers, and algorithm developers a rapid-development platform for deploying vibration condition-based monitoring.
Operationally, the development platform requires three separate units:
These elements are combined with an Intel DE10-Nano running Linux as seen below.
CN0549 hardware platform example connected to a DE10-Nano and SMA-interfaced to the MEMS sensor. Image used courtesy of Analog Devices
The kit provides the required HDL and software with open-source licensing and allows data migration to MATLAB and Python through Analog Devices’ Sensor Toolbox and pyadi-iio.
Breakdown and Features of the CN0549 System
Based on the IEPE interface, the vibration sensor technology employed by the CN0549 CBM platform uses a 2-wire standard, driven by a current source with a defined reference voltage typically between 10 V and 30 V.
Signal intelligence from the sensor is impressed on the reference voltage, modulating the signal. This modulated signal provides real-time data in both the time domain and spectral domain, which can be processed using signal processing techniques to develop coherent data algorithms.
A simplified block diagram of the CN0549 system showing the IEPE ADXL1002 MEMS sensor interfaced to the 24-bit AD7768-1 precision ADC. Image used courtesy of Analog Devices
A critical element to ensuring data integrity is centered around bonding the electronic MEMS sensor to monitor the equipment. Analog Devices has designed a cuboid mounting plate specifically to address this issue, allowing for a repeatable interface environment between the sensor and the monitored equipment.
The prefabricated cube, EVAL-XLMOUNT1, provides a high-reliability bonding interface between machine and sensor. Analog Devices recommends performing application testing on a shaker plate or equivalent. Image used courtesy of Analog Devices
The signal chain’s result is a data array of modulated voltage, quantized at 24-bit resolution. This array represents the unit-under-test’s real-time operating vibration, which can use Analog Devices’ IIO-Oscilloscope software package to see spectral content.
Example of the spectral content provided by the ADXL1002 MEMS sensor, monitoring and evaluating real-time variation based on the changing spectrum. Image used courtesy of Analog Devices
Using MATLAB or Python, data scientists can port the data stream and write analytics algorithms.
Vibration monitoring is just one methodology applied in the condition-based monitoring strategy for preventative maintenance. IoT sensor technology and Industry 4.0 automation mean less physical work for humans but an evolved skill set in systems monitoring and management.
Have you ever worked with hardware systems related to condition-based monitoring? Share your experience in the comments below.