aivis was made for industrial process data. It brings together a decade of industry experience with game-changing innovations in automated machine learning.
aivis works on unfiltered, uncleaned and unsynchronized raw data (tabular or time-series) and requires no data science expertise. It provides high quality, explainability, reliability and usability.
How is this related?
Find influencing factors, signal dependencies, and root causes for events.
Detect bad behavior
Watch critical components and systems and detect abnormal behavior in time.
What will happen?
Create soft sensors to predict lab measurements and other quantities.
React to emerging errors in time to apply appropriate countermeasures.
aivis is different
Conventional AutoML is mostly designed for business data. However, industrial process data is much larger in quantity, far more complex, primarily time-series, and mostly present in a less than optimal quality. As a result, AutoML is barely used for industrial processes leaving an incredible potential unused. aivis overcomes conventional limitations offering revolutionary new automated machine learning, mainly designed for industrial process data:
No data filtering is required. aivis figures out autonomously which part of the data is relevant for answering the question.
90% less data preparation
No data cleaning or synchronizing is required. With only one data import all use cases and settings can be executed.
No data science expertise is required. The only user input needed can be answered with domain knowledge.
No black box
Each result comes with an easily understandable report providing full explainability and actionability.
A report or model is usually created within a few minutes up to one hour. Others require several hours up to days.
aivis offers highest quality while easily handling thousands of signals and terabytes of data.
Machine Learning in Industry
The complexity of industrial processes is increasing rapidly. More and more operations data is being generated.
Machine Learning helps to understand that data and act on it. It is fundamental to increase sustainability and ensure continued control, quality and efficiency.
aivis combines a decade of industry experience with game-changing innovations in machine learning, based on in-depth research in differential geometry, stochastic processes, functional analysis, and theoretical physics.
It works exceptionally well and autonomously on process data, providing high quality, explainability, reliability and usability.
aivis helped me to truly understand the root causes of production interruptions such as sheet breaks. Instead of just curing the symptoms, which is current practice in the paper industry, we are now able to eliminate the root causes. After almost 20 years in the industry, I believe that this is a real step forward, helping me to support our customers in boosting production output.
Director Process Applications, Voith Group
With aivis we can today offer our HYDRA customers KPI analyses directly out of the box, which take the understanding of their processes to a new level.
Chief Technical Officer (CTO), MPDV
Comparison of R² of aivis and other Major AutoML Platforms (MAP) based on four different industrial datasets regarding oil viscosity, steel quality, and gas-fired power plant predictions.
The higher R², the better. As can be seen in the chart, aivis is in the lead across the board, both in terms of quality and the spread of quality.
Stay informed about the exciting news around aivis partners and products.