The desired model will allow PepsiCo to plug in ingredient attributes (protein analytical/structure/composition data, moisture content, particle size, etc.) and extrusion process conditions (pressure, temperature, dwell, screw profile) to predict finished product sensory (appearance, flavor, texture) and analytical outputs (bulk density/expansion, molecular conversion of protein, finished moisture, physical structure of).
Conversely, PepsiCo would like to be able to set desired product attributes and back out ingredient attributes & extrusion conditions to deliver those product attributes.
It may require multiple sub-models to build the full picture PepsiCo is seeking and there are expected to be heavy interactions between the long list of system inputs, which is why strong data analytics capability is key.
There may be different model systems required for different ingredients or protein structures.
Research teams who can meet all of these requirements will be preferred, but groups are encouraged to apply even if their capabilities can't accomplish this full picture.