Quick AI

Addons like Cognition were developed to allow users to create very basic AIs using their ARDI data. But there were several disadvantages to this process.

• The AIs were built on customer equipment, which may be very low-spec (particularly in embedded applications). It is usually more efficient to perform the training on better, local hardware.

• There was very little flexibility in terms of structure, size or other AI model hyperparameters.

• There was no way to add additional channels of data, drop useless channels, fix issues, filter or otherwise manipulate the input data.

• There was no way to synchronise the data with non-ARDI sources or events.

The Quick AI addon overcomes all of these difficulties by turning assets into the sample data and Python code required to build, test and debug AIs on other systems outside the ARDI server.

First, Choose the asset you want to model.

Open the ARDI dashboard for that asset, and choose Quick Model.

Next, choose the type of model you want to make.