Artificial intelligence challenges for power transformer maintenance

Artificial intelligence (AI) attempts to be involved in power transformer diagnostics are relatively old and coincide with the AI development in close domains. Artificial intelligence concept and implementation is still an emerging technology, even though the impression is that AI is ever-present. In medicine and diagnostic fields, AI has faced many challenges, and most of them are probably impossible to overcome, even by the most powerful computers. Transformer oil diagnoses and health index techniques mainly based on DGA, as well as oil tests, are some quite attractive techniques for the vital electrical industry, and they seem to be adaptable to artificial intelligence. Moreover, the combination of AI and transformer diagnosis also became attractive in the last couple of decades due to the diminishing number of human experts in this field