Answer the WHY question with aivis Insights App

Create actionable insight reports providing explainable results, that trigger “aha” moments and enable immediate efficiency improvements.

  • Simply import all your historical process data as CSV – raw, unfiltered, uncleaned, and unsynchronized
  • Use an intuitive wizard to get to the bottom of questions like Why does this happen? What is happening? or How can this be prevented?
  • Let aivis Insights analyse your data entirely on its own and create explainable reports which help you to improve your processes

You are a data analyst, domain expert, or process engineer?

Then you know the struggle to work with raw data and the pain to get optimal results out of them. This changes with aivis Insights App. It makes your life easier and gives you full control over your data.

  • Stop losing time with data preparation
    Time-series data and tabular data? Good, the only data preparation necessary is the data conversion into the aivis CSV format. Everything else aivis Insights does for you: aivis works on any amount of unprocessed, uncleaned, and unsynchronized raw data.
  • Fully focus on your questions
    Just formulate your goal with an intuitive wizard. aivis Insights figures out by itself which part of your data is relevant regarding this goal. It requires no prior exclusion of supposedly irrelevant signals or data portions nor any other insertion of pre-knowledge.
  • Improve your understanding
    Use aivis Insights to answer questions like Why does this happen?, What is happening? or How can this be prevented? providing understandable and explainable reports. Gain a deeper understanding of dependencies and hidden relationships.

Your data is secure

Every dataset we receive is strictly confidential, not saved permanently, not duplicated, remains entirely under your control, and will only be used for your purposes. Do you need an NDA? No problem.

PerfectPattern GmbH

 Three simple steps

Step 1: Upload your data

Upload your raw, unfiltered and unsynchronized historical time-series or tabular data. Both poor data quality and thousands of sensor series can be handled without any problems.

Data

time-series, tabular

Step 2: Start aivis

Let aivis automatically and independently search for hidden patterns, build models and uncover relationships, dependencies, reactions, root causes, and influencing factors.

AutoML

Step 3: Improve and benefit

Use aivis‘ insights and models to better understand, control, and improve your processes. Remove found obstacles, raise your efficiency, improve your speed, and much more.

Report / Model

Highlights

Open up whole new access to your data and realize your potential. aivis Insights is easy to use and requires no expertise in data science. Do you need more arguments? There we go:

Pragmatic

We know that you are looking for answers to real-world problems. That’s why aivis Insights minimizes the struggle to translate your question back and forth into something that mathematical approaches can solve.

Self-managed

aivis Insights requires no manual intervention. What does that mean? You don’t need to select mathematical approaches or prefiltering irrelevant data.

Powerful

aivis Insights can easily cope with thousands of parameters. This makes it especially useful in very complex environments where human comprehension quickly reaches its limits.

Explainable

To all of your questions, aivis Insights answers with understandable, explainable, and actionable reports unveiling all relevant dependencies and relationships.

Tolerant

Industrial data is not always perfect. Diverse real-world influences lower data quality with data gaps, erroneous measurements, and more. aivis Insights is exceptionally well equipped to work with any data quality.

Easy

aivis Insights works on raw, unfiltered, uncleaned, and unsynchronized raw data. Therefore, you require no other manual data preparation than CSV conversion.

Applications

quality Assurance – error detection – root cause indicators – WASTE reduction

aivis Insights provides suitable applications for variable use cases. However, the result is always an easy-to-understand report that can be easily interpreted.

Find signal dependencies

Datatype: Time-series

Find out how well one or more signals can be live predicted from the other signals within your data. The insight report will give an exact answer to that including a list of signals, that contribute to the prediction sorted by their correlation.

Check any quantity like temperature, torque or pressure but also more abstract quantities like a KPI for quality, durability or life span and much more.

Find influencing factors

Datatype: Tabular

Find out how well one or more properties can be predicted from the other properties within your data. The insight report will give an answer to that including a list of properties, that contribute to the prediction sorted by their correlation.

Check any time-independent quantity like viscosity or strength but also KPIs for quality, durability or life span and much more.

Anomaly detection

Datatype: Any

Soon available in app

Let aivis learn about the usual behavior of certain processes or components. As soon as the behavior deviates from the norm, you get an indicator about how big this deviation is.

Use anomaly detection to e.g. monitor and watch critical components that may not fail during production.

Find root causes for events

Datatype: Time-series & tabular

Soon available in app

Find and understand the root causes for a specific event. The insight report will not only provide a list of those root causes, but a classification. Each class comes with it’s own set of root causes ordered by correlation.

Classify the root causes for any type of process disruption that can be identified within the data.

Distinguish by KPI

Datatype: Tabular

Soon available in app

Identify properties, which distinguish pieces with a usual KPI value from those with an unusual KPI (bad or good). The insight report will present these properties sorted by their importance.

Identify bad apples and to figure out if those can already be identified earlier in the process.

Examples

quality Assurance – error detection – root cause indicators – WASTE reduction

Browse through our selection of different use cases, where aivis Insights has already generated huge value. Of course, those cases are only exemplary since aivis Insights is industry-agnostic and can be applied to countless other cases as well.

Paper making – Understanding paper quality

Application: Find signal dependencies

During the continuous process of papermaking, paper quality is a critical process parameter. Unfortunately, this parameter can not be measured directly because it has to be determined in the lab. Since it takes about one hour from sampling to the result, the operator is flying blind, so to speak, until the next lab value is available.

aivis Insights successfully validated that the paper quality can be predicted live from easier accessible measurements.

Dead oil – Understanding oil viscosity

Application: Finding influencing factors

The dead oil viscosity is a critical process parameter that heavily depends on the type of oil. Even slight differences in composition can dramatically impact the viscosity, making it almost impossible to address this issue with classical black oil correlations.

aivis Insights successfully validated that the viscosity can be determined directly from easily accessible measurements.

Paper making – Understanding paper breaks

Application: Find root causes for events

Paper breaks during the continuous process of industrial paper manufacturing are hard to control. These breaks can occur at any time without any warning. The sheet or paper web breaks inside the machine. Subsequently, there are various cleaning and retreading steps required before production can be continued. Sheet breaks have a significant negative impact on a paper machine’s productivity.

aivis Insights identified many yet unknown root causes, which enabled the process experts to dramatically decrease the number of paper breaks and significantly raise efficiency.

Gear boxes – Understanding critical KPIs

Application: Distinguish by KPI

Gearboxes are produced in a discrete manufacturing process. In the end, the slippage time is determined for each gearbox to check if it is within the tolerance.

aivis Insights helped to identify bad gearboxes earlier in the process by unleashing characteristics that distinguish the particularly good and bad ones from the average.

FAQ

We got your attention but you are not sure how to go on? No problem, just check out out the following frequently asked questions. If that doesn’t help just reach out to us!

Am I ready for aivis Insights?

You are ready for aivis, if you have any datasets,  time-series or tabular, that are available for export and csv conversion. And if you have a question or data analytics requirement connected to that data like why is something happening, what is happening and more.

Since the data might contain critical content or personal information, you also have to make sure, that you have the rights and are allowed to use this data with data analytics services like aivis.

Am I qualified to use aivis Insights?

You are qualified to use aivis by yourself, if you have a general understanding of your dataset and are able to convert it into the target format for aivis.

In particular, you do not need any data science expertise, since all required expertise is already built into aivis. All the questions you need to answer to start an aivis calculation, can be answered if you have a general understanding of your processes and data. The same goes for interpreting the results.

If you are not sure, just  contact our experts!


How do I make use of my data?

You have two basic options, how to make use of your data:Analyse historical data: You do this to find relationships, dependencies, root causes, influencing factors and reactions within your data that might be prone to error or just unknown so far. This might enable immediate efficiency improvements. You also use historical data to learn from past behavior and build and train models for, e.g., virtual sensors or predictions in general.

Work on live data: A trained model is a piece of software that you can integrate or deploy on-edge, on-cloud or on-site to consume live data streams. It can e.g. predict specific behavior and particular quantities or warn in time about emerging problems. Use it to ensure quality, increase safety, raise throughput, and much more.

What kind of data do I need?

There are two basic types of datasets you can use aivis on: Time-series data and tabular data.

Time-series data consists of data points, that usually come from a sensor or signal and are associated with a timestamp that determines when exactly the value was set or received. Read more about time-series data on  wikipedia. See also the aivis  time-series data specifications.

Tabular data entries are usually created when an event or state is logged. Although it might contain timestamps, it can also be completely time-independent. Compared to time-series data, where a value is taken e.g. every second, tabular data is more event or state oriented. It might also describe not a point in time, but time frames. See also the aivis  tabular data specifications.

The amount of data can be anything between a few kilo bytes (kb) and many terra bytes (tb).

How to prepare my data?

Whether you have time-series or tabular data, aivis works best with raw data as little processed as possible. Especially no records should be removed or interpolated. Time-series data does not have to be synchronized, nor is it necessary that the timestamps are equally distributed.

You need to convert this data to CSV format following our  specifications for time-series and tabular data. If you do not know how to do that,  contact our experts.

Data transfer: You can either upload the data with the  aivis Insights app or store it in your own Amazon AWS S3 storage and create a user with privileges to read the dataset. For larger data volumes, it is possible to send hard drives with the data to us.

Will my data stay under my control?

Yes! Every dataset uploaded to the aivis Insights app is strictly confidential, not saved permanently, not duplicated, remains fully under your control and will be used for your own purposes only. An NDA can be set into place.

Want to unlock your data with the aivis Insights App?

Fill out the form to contact our experts