After years of experiencing first hand the numerous challanges of collecting machine condition data from industrial plants, Ascenix is now at the latest stages of introducing a reliable, affordable and scalable platform to stream sensor waveform data right into machine health prognostic applications. Bringing together experts in the fields of predictive maintenance, signal processing,artificial intelligence, as well as senior software and hardware engineers as founding partners, Ascenix is equipped to become one of the innovators in sensor data streaming in the IoT age.
The demand for more autonomous plants, i.e. Industry 4.0, spurs the research and development in machinery health prognostics as machine useful life estimation and predicting machine failures is an integral part of the autonomous decision making process which aims at 'near-zero downtime' performance.
One of the main challenges that remain in this field is the acquisition of the high speed waveform data from the industrial settings for the core vibration and current signal analysis, although well-defined methodologies and failure signatures are at hand,
Bigger burdens on the dataset, e.g. full, months long run-to-failure period data, 20K+ samples per second streamed from a single sensor, etc., are being brought by the newer predictive analytics, e.g. articificial neural networks.
A dedicated open platform is highly essential, as automation platforms are already at their capacity and proprietary condition monitoring systems or other factory instrumentation can only share limited raw data, if at all.
Ascenix iProbe platform features IoT enabled, fast, multi-channel data acquisition units and fog computing architecture with powerful edge computers and AWS S3 storage.
Sensor data streaming efficiency, key for scalability, is managed by novel compression and transformation tools running on the edge computers, achieving over hundred fold data size reduction.
Sensor data streaming reliability is maintained by the TCP/IP sockets implementation on the daq units, and packet loss and inter-packet time difference is tracked rigorously on the driver side.
Timestamp accuracy, also critical for data enrichment, is maintained at the daq unit level and sub-ms resolution is achieved. Encoder/tacho triggered data acquisition can also enabled.
Affordability, so that engineers can initiate their own projects on an expense budget or hundreds of measurement points can feasibly be included, is made possible by a compact, no-frills, data acquisition unit with both signal front-end and communication modules, and a Software-as-Service Platform on the cloud.
How it works
Smart and Simple 1-2-3
Step 1: Connect your sensors, provide a LAN connection and the the Ascenix ADCR daq device will find the edge computer and reigster itself.
Step 2: View ADCR status and send the initial config on the control dashboard running on your server or laptop.
Step 3: Point your prognostic application to the ADCR's cloud storage and you're ready to go. Use existing data sets on demand, without any new capture activity.
Fast and Reliable
Sustaining up to 64 Khz per channel data on 8 channels simultanously and continously, ADCR's fulfill the requirements of any streaming waveform capture implementation.
Accurate timestamps and sustained continuity on the data stream provides for a relaible and useful dataset for any predicitve analytics.