Empower Agent v5.1.x: CDS data integration at enterprise scale

October 20, 2023

TetraScience is happy to announce the release of our newest Empower Agent, version 5.1.x. The Empower v5.1.x agent is the result of a rigorous design and development process spanning multiple quarters, including real-world, live-instrument testing. This release features five significant improvements:

  1. Balances the stability and latency to achieve better predictability 
  2. Enables lab automation use cases with new bidirectional commands for sample set method creation
  3. Provides tunable parameters that unlock proven vertical scaling
  4. Offers a solution for the Empower SDK limitation for mass spectrometry data
  5. Allows users to select multiple projects when rerunning projects

These features support enterprise-scale deployments and Empower integration with thousands of workflows.

1. Balancing stability and latency to achieve better predictability 

One of the key requirements for scientists using the Tetra Scientific Data CloudTM is the ability to upload data from Empower as soon as the injection is ready. Conceptually, this is a straightforward process. However, in practice, it was never as simple as that. 

First, when an injection was detected by the Tetra Empower Agent, it did not indicate that the data was ready, as ingesting the data immediately could have led to partial data extractions. 

Second, since Empower is an enterprise-level system, it can hold thousands of projects from as many scientists, at any point of time. Additionally, these projects contain multiple types of activities including: 

  • New injections performed by scientists 
  • Historical projects restored from NuGenesis or the Tetra Scientific Data Cloud

With this project volume and complexity, the timing of uploads for newly changed injections were often unpredictable.

To address these challenges, TetraScience is introducing two major changes: 

  1. The system can now identify high-priority projects based on recent modifications. The new feature includes a configurable setting that helps ensure there is a dedicated process to detect injection changes. This facilitates the quick and efficient ingestion of newly created or modified injections, without users having to wait for historical updates to process during migrations.
  2. The Empower Agent now provides a configurable batch size for the scanning process. When there is an unhandled error by Empower ToolKit, such as 0x80010105, the Empower Agent can now recover more predictably and proceed to the next set of working projects. Users can choose a batch size of 1 to 5 to introduce more predictable latency for change detection. This reduces downtime and helps enterprise-level data workflows flow more predictably.

2. Enabling lab automation use cases with new bidirectional commands for sample set method creation

To further enable scientific use cases such as batch release and stability analysis, 5.1.x agents now feature commands to support “Creation of Sample Set Method” and “Fetch Sample Set Method.” 

This feature has been thoroughly field tested and is now available for all users. It allows customers to create a sample set method from scratch, or from an existing method, as a template through the “Fetch Sample Set Method” command. These bidirectional commands greatly reduce the time spent on method testing for scientists.

3. Unlocking vertical scaling

At the same time, with Empower Agent v5.1.x, TetraScience added various tunable parameters. Users can accelerate throughput through deeper vertical scaling, employing a more powerful VM. The following benchmarking test demonstrates that by vertically scaling the Empower Agent VM from m5.xlarge to 2xLarge, and tuning the agent parameters, the agent can nearly triple its throughput speed.

4. Solving the limitation for mass spectrometry data

Due to a Waters toolkit bug in previous agent versions, the data cube data returned from the “Get3D“ method loses precision when fetching data from the “3D MS” channel. The “GetSpectra” method suggested by Waters was slow because it had to fetch data at every single retention date time point, instead of using one method call. Based on our testing, on average, it takes 100 times longer to fetch the same injection by using “GetSpectra” compared with “Get3D.” In this release, we provide users the ability to configure the underlying method on a project level.

5. Multi-project selection for “Rerun Projects”

In previous versions of the Empower Agent, when users wanted to execute a “Project Rerun” to retrieve new or updated data, users had to select and “rerun” projects one at a time. If a customer needed to rerun a large number of projects, this process could be time consuming.

The Empower v5.1.x provides users with a Rerun Project selection list. Users can now select multiple projects to rerun from a single screen, greatly reducing the number of clicks required to rerun project sets.

Rerun Projects tab in the Empower Agent v5.1.x with multiple projects selected.

Improved enterprise-level CDS data management with Empower Agent v5.1.x

The Empower Agent v5.1.x represents one of our most exciting agent releases to date. The improved predictability, bidirectional command capabilities, accelerated throughput speed, mass spec data configuration, and multi-project selection functionality for project reruns will significantly improve both product performance and user experience.

If you have any questions about this release, or you would like to see a demonstration of the Empower Agent v5.1.x, please reach out to your TetraScience account team and they can set up a meeting with a member of our Tetra Integration team.