Methods and apparatus for obtaining diagnostic information, methods and apparatus for controlling an industrial process

A method of obtaining diagnostic information relating to an industrial process, the method comprising:

• implementing a first sub-domain implemented as electronic instructions and/or data in one or more non-transitory computer-readable media, the first sub-domain being or corresponding to a model representing at least part of the industrial process and comprising first diagnostic information extracted by analysis of first object data, the first object data representing one or more first parameters measured in relation to a first set of product units that have been subjected nominally to the same industrial process as one another;
• implementing a second sub-domain implemented as electronic instructions and/or data in one or more non-transitory computer-readable media, the second sub-domain being or corresponding to a model representing at least part of the industrial process and comprising second diagnostic information extracted by analysis of second object data, the second object data representing one or more second parameters measured in relation to a second set of product units that have been subjected nominally to the same industrial process as one another;
• and
• implementing a network comprising at least one probabilistic connection from a first variable in the first diagnostic sub-domain to a second variable in the second diagnostic sub-domain, wherein at least some information of, or used to extract, the second diagnostic information is influenced probabilistically by at least some information of, or used to extract, the first diagnostic information.

A non-transitory computer program product comprising machine readable instructions therein, the instructions, when executed by a data processing system, configured to cause the data processing system to at least:

• implement a first sub-domain as electronic instructions and/or data in one or more non-transitory computer-readable media, the first sub-domain being or corresponding to a model representing at least part of an industrial process and comprising first diagnostic information extracted by analysis of first object data, the first object data representing one or more first parameters measured in relation to a first set of product units that have been subjected nominally to the same industrial process as one another;
• implement a second sub-domain as electronic instructions and/or data in one or more non-transitory computer-readable media, the second sub-domain being or corresponding to a model representing at least part of an industrial process and comprising second diagnostic information extracted by analysis of second object data, the second object data representing one or more second parameters measured in relation to a second set of product units that have been subjected nominally to the same industrial process as one another;
• and
• implement a network comprising at least one probabilistic connection from a first variable in the first diagnostic sub-domain to a second variable in the second diagnostic sub-domain, wherein at least some information of, or used to extract, the second diagnostic information is influenced probabilistically by at least some information of, or used to extract, the first diagnostic information.

A diagnostic method for an industrial process, the method comprising:

• implementing a first sub-domain as electronic instructions and/or data in one or more non-transitory computer-readable media, the first sub-domain being or corresponding to a model representing at least part of the industrial process and comprising first diagnostic information associating probabilistic first context data to a first product unit based on one or more first parameters measured in relation to the first product unit;
• implementing a second sub-domain distinct from the first sub-domain, the second sub-domain implemented as electronic instructions and/or data in one or more non-transitory computer-readable media, being or corresponding to a model representing at least part of the industrial process, and comprising second diagnostic information associating probabilistic second context data to a second product unit based on one or more second parameters measured in relation to the second product unit;
• and
• implementing a probabilistic connection between the first context data and the second product unit.