A new way of working

– make the most of your data and your competence

At Torch's core of is a powerful analysis engine. The engine consists in a data flow graph for which the nodes constitute computational units, such as filters. In a very transparent way, input data is gradually transformed into business critical information, as it flows through the graph.

 

Torch provides a powerful graphical interface to the analysis engine, called the Atomic View. In this view, the complete data flow graph behind each analysis is visualized. In addition to providing a very clear idea of what is going on, it allows for a unique experience of interactivity.

BE PREPARED FOR TOMORROW

Configurable

The user interface is very flexible and may be configured differently for different types of users.

Transparent

The exact way all analyses work is fully exposed to the user

Efficient

Based on powerful caching and parallell computing paradigms

Traceable

The underlying data flow graph makes it possible to follow any data, from input to end result.

Scalable

Designed to handle tomorrow’s amounts of data 

Sheringham Shoal Offshore Wind Farm - Statkraft

Search for root causes

The Atomic view provides support for

• Adding a data view to any edge in the graph to check intermediate results.

• Changing and playing with parameter values to see how the results are affected

• Editing the Python code for specific nodes and see the effect on the results in the views

• Excluding certain parts of the computations – skip/bypass certain nodes

• Changing the order of operations

Tailor the built-in functionality

How the built-in analyses work is clear from the Atomic View and thus completely transparent to the user. If they don’t work as expected, it is easy to do modifications so that they fit your needs:

• Adding new nodes to an analysis to introduce more complexity

• Changing the order of nodes

• Changing parameter values

It is very easy to add new analyses to the system. You don’t need to wait for the next version of the software to get a new feature. You just do it yourself from the Atomic View by

• Creating a new data flow graph. Either start-from scratch or use one of the built-in analyses as a template.

• There is support for adding, removing and connecting nodes.

• Python scripts are easily associated with the nodes

Expand the feature set of the system