Predictive Maintenance

Our AI-based Predictive Maintenance monitors the condition of your machines to predict and avoid downtime!

Our webinar on "Predictive Maintenance"

Review our webinar on Predictive Maintenance and how you can increase your asset productivity using Artificial Intelligence (AI) methods. Predictive maintenance is one of the major application areas of artificial intelligence in the context of Industry 4.0. The video starts playing after you entered your data.

Our solution

Our Predictive Maintenance solution uses Artificial Intelligence and Machine Learning to prevent equipment failures and unplanned downtime.

90 % higher productivity of service personnel

80 % prevention of unexpected failures

40 % less inventory and waste

30 % longer plant life

Our approach

By examining historical sensor readings, permanent condition monitoring, and analyzing real-time data, the machine learning algorithm can learn the relationship between sensor readings/changes and historical failures to predict similar failures in the future, contributing to a more effective maintenance strategy.

One of the approaches used to predict the remaining machine life before failure is regression modeling algorithms.

We also use classification modeling algorithms as one of the ways to predict in which time period the failure will occur and what type of failure will occur.

Our process

Data collection

  • Connection with various data sources
  • Harmonize to common data storage or stream flow

Data pre-processing

  • Integration of the data sources
  • Resampling to common time base
  • Handling gaps in the data
  • Dimensionality reduction

Dataset preparation

  • Dealing with dataset imbalance
  • Data standardization
  • Preparing training and test data

Production deployment

  • Continuous learning of new events
  • Monitoring for unexpected
    performance deterioration
  • Reporting alerts and model results

Model optimization

  • Testing performance
    on all provided data
  • Automatic parameter tuning

Model selection

  • Task specific model selection
  • Cross-evaluation against baseline
  • Benchmarking of the proposed solutions on selected data

You want to learn more about Predictive Maintenance?


KENDAXA is an Artificial Intelligence software solutions company that makes people unburdened and insightful, so as to make them more efficient, not redundant. We are expanding the market for sophisticated AI, Machine Learning and Data Sciences in manufacturing and financial services, and make it affordable for mid-market customers by delivering better, faster, cheaper AI that we call Intelligent Automation. We provide advanced, configurable, repeatable, scalable, SaaS-deployed AI-Apps – which is what people want – powered by an innovative platform and industry expertise. We’ve been at it since 2017 and have great people in Leipzig, Germany and Prague, Czech Republic and a few other places.

Our customers

Automotive Dealer

Insurance Broker

Investment Banking

Building Materials

Receivables Management

Automotive Banking

Aviation Supplier