We are thrilled to announce the release of the Event Autotagging feature. Using customizable configurations, Model-Prime’s Event Autotagging accurately identifies and categorizes important events within log files, saving time and effort when compared to manually reviewing log files for events. This feature adds value to teams needing to identify important occurrences in large volumes of log data.
The Need for Event Autotagging
We built this feature after hearing many customers ask for a way to tag events of interest automatically. Data overload has long been a significant barrier to gaining meaningful insights from large-scale robotics data. Robots are relentless data generators, producing vast amounts of information with every operation. As an organization grows, the volume of data only multiplies, and manual inspection of logs becomes inefficient.
Benefits of Event Autotagging
Event Autotagging addresses this challenge head-on, offering robotics companies a valuable tool to:
- Build higher quality machine learning models faster by training on tagged events.
- Reduce the need for manual inspection of logs by automatically identifying and tagging events of interest.
- Monitor and analyze critical events for performance analysis by pinpointing areas for improvement.
- Enrich log data with event-related attributes to facilitate searchability and data-driven decisions.
- Support needed reporting for events relevant to safety or regulatory needs.
The Technology Behind Event Autotagging
At the core of Model-Prime’s Event Autotagging feature is a powerful yet user-friendly tagging system that caters to both coding enthusiasts and those preferring a low-code approach. Developers can use the flexible and extensible framework to define custom taggers in the Model-Prime web UI or through the API.
How it works
- Using our low-code editor, configure a tagger within the Model-Prime web UI using a set of Python-based conditions for scanning log files. These conditions specify the required input channel data, event criteria, and outputs in the form of metadata attributes.
- Model-Prime scans through log files, looking for continuous blocks of frames that match the defined conditions.
- When a matching block is found, Event Autotagging automatically tags the relevant time range in the log with the specified metadata attribute. Once created, the attributes associated with each event are readily available for further analysis, visualization, search, and integration with Model-Prime tools and workflows.
Integration and Compatibility
Recognizing the diverse ecosystem of robotics platforms and data formats, Model-Prime has designed Event Autotagging with integration in mind. Out of the box, the feature integrates with ROS 1 and ROS 2 bag formats, ensuring a smooth transition for companies already utilizing these widely adopted robotics frameworks.
Beyond these standard formats, Event Autotagging has a robust and well-documented API, enabling companies to integrate this feature into existing workflows.
The Future of Event Autotagging
Event Autotagging is another step towards helping robotics companies get the most out of their data. Model-Prime is committed to continuously enhancing and expanding this feature. We are driven by a deep understanding of customer needs and feedback, as well as the ever-evolving landscape of robotics and data analytics.
Schedule a demo with Model-Prime today to learn more about Event Autotagging, and stay tuned to learn about future feature iterations.
