Concealed within exabytes of sensor data generated by industrial machines are micro-patterns that can tell us when a machine is likely to fail. Until now, these patterns could not be recognized, even by the most advanced statistical packages.
Presenso develops solutions for Predictive Maintenance within the Industrial Internet of Things (IIoT). It presents this information directly to maintenance and reliability professionals so its clients no longer need to hire Big Data experts to deploy and maintain AI based solutions.
Presenso continuously streams asset sensor data to the cloud where Artificial Intelligence algorithms analyze it in real time. The platform is sensor-agnostic and can monitor signal data without the need for manual human input like the setting of control limits.
A Wide Array of Services and Solutions
Presenso’s Cloud-based software solution replaces the rules-based legacy systems which cost manufacturers and plant operators millions of dollars a year. Reactive in nature and with limited computational power, those outdated industrial monitoring tools are unable to effectively control production downtime.
The revolutionary cognitive software from Presenso provides unparalleled operational intelligence and deep semantic insights which increase production yield and revenues. Its competitive range of solutions eliminate manual intervention and the need for expert knowledge.
They are hardware-agnostic, can be rapidly deployed in remote locations, and incorporate deep learning capabilities within the analytics engine. By utilizing the latest machine learning and Big Data technology, they add value to anomaly detection by improving correlation, prediction and prescription capabilities.
Advanced Deep learning and Machine Learning algorithms analyze asset sensor behavior and automatically detect abnormalities and patterns within them. After the detection of anomalies within the signals, correlations and pattern detections are analyzed automatically. This information and the exact sequence of abnormal events can then be presented to operators.
Once an evolving failure has been detected, a failure alert is generated. This alert includes information on correlated sensor abnormalities. This valuable information is essential to tracking the origin of the failure.
Presenso’s industry-agnostic IIoT predictive maintenance has seen diverse adoption across many fields:
- Power and Energy:
Presenso’s customers receive machine failure analysis and predictions from the entire power plant, from single, small turbines to fleets of large-capacity turbines spread across multiple power plants.
- Oil and Gas:
Presenso’s solution improves production continuity and increases Overall Equipment Efficiency (OEE) across all fields and facilities.
- Water Facilities:
The company’s solutions aid water desalination facilities and waste water treatment facilities to avoid downtime and meet the ever-growing demand for drinking and agricultural water.
- Automotive Industries:
Presenso’s customers can analyze manufacturing floor data in real time, get a clear overview of the performance of assembly machines, and utilize condition-based maintenance.
When the company was founded in 2015, the market was far less receptive to Machine Learning for IIoT. However, over the last 18 months, Presenso’s proven performance and stunning success across multiple market verticals has led to a surge in interest and adoption.
Presenso’s Automated Machine Learning is based on innovations in Artificial Intelligence that were previously not applied to Predictive Maintenance.
One of the greatest challenges to deploying Industrial Analytics is the need to select the right Machine Learning algorithm for a given dataset. Presenso’s Auto ML eliminates this difficulty with a library of hundreds of algorithms that can be used. The system itself selects the optimal algorithm for the data without the need for human input.
Another advantage is the Unsupervised Machine Learning methodology that the company applies to Big Data. The algorithm is independent of sensor, vendor, asset, age, machine, and process. It can automatically identify data anomalies without the need to first learn the underlying process it is monitoring. This is an important differentiator because it allows for rapid and relatively inexpensive deployment of the Predictive Maintenance solution.
Eitan Vesely is the Co–Founder and CEO of Presenso. He is a mechanical engineer by education.
Eitan previously worked as a systems support engineer at Applied Materials, where a major part of his job was troubleshooting manufacturing plant hardware failures that led to machine shutdowns. Another aspect was working through reams of data before traveling to customers’ plants to bring failed machines back into production.
Deddy Lavid, Presenso’s Co-founder and CTO, has previously worked on energy consumption prediction. With a strong background in AI and predictive analytics, he extended those concepts to a full-scale Smart Factory with advanced analytics to scrutinize all sorts of industrial data from various arms in the manufacturing value chain.
Dr. David Almagor, the third Co-founder as well as the Chairman of the company, boasts over 30 years of experience managing complicated research and development processes as well as business entities. The serial entrepreneur has led various start-ups towards success with his leadership abilities and unwavering attitude and hopes to continue the same with Presenso.
Adaptability and Foresight
It is generally accepted that the typical Maintenance and Reliability engineer cannot perform the role of a data scientist, nor can he or she be the bottleneck for deployment.
Companies such as Presenso recognize that it needs to continually innovate on AI capabilities but make the front end user experience as simple as possible. Applying AI to sensor-generated Big Data is not scalable in today’s industrial plants unless the analysis is performed using automated Machine Learning tools.
Presenso is investing significantly in R&D in order to reduce the time it takes for algorithms to make a prediction. Machine Learning for Automated Algorithm (Auto ML) is only the first step in that journey.
The company’s goal is to simplify IIoT Predictive Maintenance for the end-user so an industrial plant can be alerted to upcoming failure with enough time to avert the problem.