Product User Community

Cognite Data Fusion is a product built to address the challenges of working with industrial data by: Making data available - Liberate their IT, OT, ET and visual data from siloed source systems with our extractor pipelines. This is done reliably and at scale. Making data meaningful - We use AI-powered contextualisation services to create an Industrial Knowledge Graph that delivers trusted, contextualised data Make data valuable - Cognite Data Fusion enables your teams to access this data with the best-of-breed tools of your choice to turn this data into business value Monolith solutions often end up creating vendor lock-in and can even end up creating more data silos within your organisation With trusted, contextualised data available in an industrial knowledge graph, your teams are equipped to scale solutions both in the volume of new solutions and replicating successful solutions across assets, lines, or sites.The videos are based on the “ice cream factory” use case: a use case

 Hello Charts Early Adopter Community,There have been several new features and functionalities released lately, with some being released just today. I’ve recorded a video (below) to explain several of these core features in detail, namely – calculations running on individual data points, not aggregates.You can also scroll down to read the written details.Please do leave comments below with questions and feedback!DetailsCharts UI/UX Calculations are now running on individual data points, not aggregates In the past, calculations have been running on time series aggregates. This meant that, although calculations were approximately calculations were often not correct his is a very important new feature that greatly increases the accuracy and trustworthiness of calculation results in Charts. Watch the above video or see the slides, below, for more info. If you receive a warning that looks like this (see image, below), then it means that the results of the data has been downsampled to perfo

WhyWhen starting out in Cognite Data Fusion (CDF) project, it's natural to start by creating data governance elements like groups, datasets, and Raw databases from the CDF user interface. But as the solution begins to scale, you'll quickly realize that it is demanding to set up a detailed configuration handling multiple solutions, sources, roles, and other dimensions. For scaling, precise control is needed for access management and data governance and to enforce the guidelines and rules across the solution. The problem is not trivial, and a good way to solve it is by replacing the manual approach with a configuration-driven system, where the configuration language supports higher-level concepts for data lineage and access control. With configuration files as the foundation, you can set up an automated DevOps process and review and approve any changes to the structure before they are deployed. This approach also dramatically simplifies sharing the same configuration across multiple envi

Just stumbled upon an article from McKinsey in 2019 which articulates well what I believe is a crucial role at many of the customers I’ve worked with: the Analytics Translator. The technology in industry is continuing to move forward quickly opening up many new opportunities, but the deep domain expertise in production processes, maintenance, and more is still as important as ever. To me, the Analytics Translator role seems a perfect bridge between these two worlds. I’ve seen a number of manufacturers successfully “upskill” prior machine operators, process engineers, and more to fulfill this role. It’s always very impactful, making for significantly more effective digital initiatives. From my personal experience, I see them doing things like:Running use case workshops and prioritizations Internal training of users Adoption tracking Interim project manager during new technology initiatives Liaison with external technology partners and SIs Curious how others see this role, and whether yo

Cognite Data Fusion is a product built to address the challenges of working with industrial data by: Making data available - Liberate their IT, OT, ET and visual data from siloed source systems with our extractor pipelines. This is done reliably and at scale. Making data meaningful - We use AI-powered contextualisation services to create an Industrial Knowledge Graph that delivers trusted, contextualised data Make data valuable - Congite Data Fusion enables your teams to access this data with the best-of-breed tools of your choice to turn this data into business value Monolith solutions often end up creating vendor lock-in and can even end up creating more data silos within your organisation With trusted, contextualized data available in an industrial knowledge graph, your teams are equipped to scale solutions both in the volume of new solutions and replicating successful solutions across assets, lines, or sites.The videos are based on the “ice cream factory” use case: a use case

Today’s topic is about interacting with Cognite Data Fusion (CDF) from Android devices. Android devices can be phones, but also other kind of devices like hands-free devices (with speech commands). Since Kotlin, the Android preferred language, has interoperability with Java, it facilitates the integration of interactions with CDF in your Android apps. You can indeed use our Java SDK in an Android application, which makes it really easy to retrieve and upload data from/to CDF. The Cognite Java SDK is complete and maintained. This can broaden your thoughts/ideas about what to do with CDF. Below a few examples of what can be done : Retrieve information from CDF to display it on your mobile device Upload data to CDF from a mobile device, which can be useful in an industrial context Using specific features, from mobile devices (like a camera for example) to upload particular data to CDF (raw, or to extract features from them with an ML model, thanks to a Cognite function for example)   

Hi, Just wanted to know if there's any way to use Cognite UI Apps instead of we developing custom UI App. If yes, could you please let us know the process and how to use it? 

We have identified a bug in the time series datapoints fetcher in the Python SDK which may cause duplicated timestamps to be returned. Affected versions are 2.47 through 2.51. Please check if you have anything deployed using the datapoints API in this version range. If you do, please upgrade to >=2.52 ASAP.https://github.com/cognitedata/cognite-sdk-python/blob/master/CHANGELOG.md

This is how it looks like when you push geospatial to the limit and show 4+ billion essential ocean variables (temperature, salinity, pH, oxygen etc) captured from 220,000 research cruises from 1890-YTD. Such data is very important to understand the effects of climate change, for instance ocean warming, deacidification, dead-zones (lack of oxygen), biodiversity migration and more. In the Ocean Data Platform, we have implemented something called the “Ocean Data Connector” which is a cloud-based JupyterHub infrastructure where you can analyze this data in a very efficient way.Ocean Data Platform and the “Explorer” interface 

On 8th March our amazing data scientists at HUB Ocean and Cognite gave a workshop at the Woman in Data Science (WiDS) event in Oslo. Have a look at the video and blog post. The data challenge was related to coral reef bleaching events applying open data on the Ocean Data Platform.Closing the gender gap in Ocean Science? — HUB Ocean | Dedicated to Unlocking Ocean Datahttps://www.linkedin.com/posts/hubocean_coralreefs-biodiversity-collaboration-activity-6928641478181154816-IImk?utm_source=linkedin_share&utm_medium=member_desktop_web

As we announced in our last release post, we recently updated our calculations backend to run on individual data points if the total count of data points is less than the predefined maximum limit.(You can watch the video walkthrough for more information.)When this improved functionality was released a few weeks ago, this limit was set at the intentionally low value of 10,000 individual data points. We did this to test our infrastructure, gather feedback, and ensure our backend will not crash with these more expensive requests — Thanks to those of you who have provided us with input!As of today, we have released an update which increases this limit from 10,000 data points → 100,000 data points.This 10x improvement in performance will help to provide accurate, trustworthy calculation results for larger ranges of time and data. With this new 100k limit, we’ve reached the maximum number of data points we can retrieve from the Cognite Data Fusion Time series API with a single request.While

 Hello  @brendan Buckbee at Celanese has created a chart for the 12 month, but gets an error message. Is there any way that this can be created into a chart. I have attached a screenshot of the calcs. cc: @Kylie R  

Putting your learning into practice can be challenging. Not to worry, we're here to help! 🚀  Our subject matter experts and the Cognite Academy team joined forces to create a course on CDF Transformations. The course is designed for data engineers and anyone who wants to learn about CDF Transformations.In this course, you'll learn how to transform data into the CDF data model using CDF Transformations. This course will walk you through lessons on the target schema, writing SQL queries, and running and scheduling transformations, accompanied by interactive knowledge checks, reading materials, and a hands-on exercise.Upon completion, you'll be able to: Understand why you should use CDF Transformations.  Find detailed information about the target schema. Write SQL queries and use SparkSQL and Cognite's custom functions to transform data. Schedule and run data transformations. Register for this course right away! 🕰Happy Learning! 😊

Found a bug? Have a question about how something works? Want something new with Flexible Data Modelling? We want to hear about it!You can choose to either create a dedicated post (topic) in the Flexible Data Modelling group by clicking the Create topic button OR simply post a reply below in this thread.Be sure to tag with “bugs” or “feature request” Features requests You can choose to either create a dedicated post (topic) in the Charts group by clicking the Create topic button OR simply post a reply below in this thread. Remember to say whether your feature request is Nice to have, Important, or Critical to you and why. Screenshots, sketches, or explanatory videos are also encouraged. We will follow up and share progress on features periodically as well! Bugs Remember to include a screenshot or video to help the product team best understand what exactly you’re talking about or referring to.

Hi, In the AIR application, one can choose to see the relevant time series for 5 years, 1 year, 6 months, 1 month, and “Custom”. When clicking “Custom”, a date choosing panel shows up. When selecting the end date, the application freezes, and one has to refresh the application to get it back. (Feilkode: RESULT_CODE_HUNG)  

Hi. How can I get aggregated values from synthetic timeseries fields on templates? Intuitively, I would expect the “datapointsWithGranularity” operator in the graphQL query to give me this, but I get the same results as for the “datapoints” operator. Should the aggregation type and granularity be part of the definition of the synthetic timeseries field in the template? If so, how? 

Hi! I’m curious what is the best practice to do this basic thing and visualize on Grafana ( ideally without code): With a flow transmitter, summarize the total volume per day or week.  Correct the above against a threshold or a running signal. Ie. I now when the flow should be 0Here’s how I do nr 1, it feels a bit like a hack. Nr2 I am solving with Functions in CDF, however it would be nice if there’s a quick way of doing it without code,  The query looks like this: 24*ts{externalId="arba:one:s=0:PLC_03!Vc_S07_200_FT01_PV", alignment=1653386400000}-Should the alignment be at the middle of the day, 12:00, or at midnight 00:00? When we get back to wintertime, it will be off by one hour..  

Hi, When parsing a large production model some high level concepts we want to filter on are structured as assets. I.e. a fiscal region for power in the Statnett case.  Our model is now hitting limitations of subtree queries. With more than 100k assets per region for some regions. What is your thinking around how to handle such cases in a model? To spike the conversation we’ve considered moving high-concept parts of the tree to labels, or making more of the DB type of operations locally. However, the former demands a pipeline for moving concepts suited for a tree  to a Label “just because”. The latter requires quite a lot of iron present on the local instance processing the query. An instance of the SDK quering we do can be seen in the power-SDK in github.

Hello, Today’s topic is about Cognite functions: Cognite functions enable the user to run and schedule arbitrary code to clean, process and do calculations on your data. They are easy to use and have seamless integration with Cognite Data Fusion (CDF), allowing the Cognite Function to interact with CDF directly. The service automatically scales the computing infrastructure to handle a fluctuating number of function calls.   They can be triggered on schedule, on click in the UI, and even with the experimental Python SDK, which makes them adaptable to a lot of use cases. We also have a full CI/CD flow for them, which allows to deploy and maintain them really easily. Some examples of use cases that can be done with Cognite functions:Change point detection, thanks to machine learning models, that run on schedule Anonymisation of videos, with computer vision models @Gaetan Helness is quite familiar with Cognite Functions. If you have any questions, don't hesitate to reach out to us in t

After confirmation from @Geir Engdahl, is there any plans for developing models where application developers can attach data to users, like preferences or even work orders?

The best way to play around with Flexible Data Modeling in CDF is through Templates.  Please be aware of the following before getting started! Currently, the App Dev Journey team at Cognite is working on a new Schema Service that will replace Templates towards the end of 2022. For more details on the differences between Schema Service and Templates, go here.There will need to be a migration between Templates and Schema Service in the second half of 2022. The migration will be simple for the data modeling features and querying features as they will be both be using a GraphQL interface. However, moving the data could be a more difficult migration.Hence we highly recommend continuing with Templates for POCs and early experimentation. You can also productionalize with Templates, but please reach out further clarification around quota and SLAs on this Cognite Hub group. Click to see the official documentation on Getting Started with Templates

 Hello everyone! Nicklas from InField here 😊 It’s been about a month since we were on the grounds with the crew at Celanese, and the InField team has been hard at work analyzing, prioritizing and planning work going forward based on all the insights gained from the visit. The visit has been invaluable and we want to keep the momentum going full steam!While we’re working on delivering new functionality and fixing bugs, I just wanted to reach out and say a huge thank you to the awesome and helpful team at Acid South, that makes this possible – we learned so much from you and it was a blast to work together on the grounds and I am very hyped about our collaboration going forward! 😊 See y’all again soon, Nicklas on behalf of InField   Cc. @Crystal Connor Richards @Kylie R @ibrahim.alsyed 

Introduction In this post we continue to share some of our internal material aimed towards solution builders, such as data scientists, who want to develop their ability to develop high quality solutions by creating more reliable, maintainable and readable code. This is the second part of 2.MotivationHigh code quality is easy to recognize but can be very hard to describe concretely. The assumed benefits are easier maintainability, modifiability, and more. While code style, like formatting, can be a matter of different taste, most parties agree that other code practices that fall under the umbrella term “anti-patterns” should be avoided. To stop endless formatting discussions and the like, having (and adhering to the industry) standard makes reading and understanding code across repositories easier.What this guide is notThis guide will not tackle the topic of  “how to set up a Python project” the right way . Please let us know in the comments if you would like us to share more of our ex

The Cognite OPC UA extractor reads data from OPC UA servers and writes data to Cognite RAW, assets, time series, events, and relationships, depending on how it is configured.  Now, we’ve made the GitHub repository public. The extractor serves as a comprehensive example for using the .NET extractor utils, or as a baseline for developing custom applications working with OPC UA and CDF. The repository also contains the Cognite OPC UA Extractor test server, which comes with a CLI and can be used to test OPC UA client applications. Use this test server to simulate common server issues and generate data points, events, and even new nodes and references. With the open-source OPC UA extractor code, you’ll get insight into how Cognite works with OPC UA which will hopefully make it easier to develop your own extractors.

I am doing some ETL jobs in Azure Databricks and have successfully managed to use Cognite’s Spark Data Source to read and write time-series, datapoints etc from and to CDF. I know that databricks itself is a cloud platform. However, it is interesting for me to be able to run some or all of the jobs locally during development phase. I wonder if it is still possible to somehow test-run Spark jobs locally? The configuration does not seem to be trivial.  I played a little bit with PySpark, and I was able to run it on my Mac but I could not create a connection to “cognite.spark.v1” to read or write data. Do you know if it is possible to perform such operation? If not, what would you suggest?