© D.R.
In our previous article, “The Augmented Worker: What AI will change for technical auditors“, we saw how visual recognition can create virtual assistants that can increase the productivity of operational staff, the quality of their services, and the value of their work. Today, we will see how the integration of AI into operational businesses is also a great opportunity to transmit and structure business expertise, regardless of the type of process to be optimized.
Digitizing the activity
Let us take again the example of the optimization of the management of water and sanitation networks through the detection of anomalies of all kinds. Several factors can cause a leak. Some are simple to detect (roots in pipes, presence of rodents), but others are much more subtle and require the intervention of an expert with solid skills acquired over many years. This is particularly the case for the diagnosis of the different categories of corrosion, which are difficult to identify and therefore to treat.
To train an AI to recognize the wide variety of potential problems and become a virtual assistant, it is first necessary to build a complete, explained and documented data set in order to create a good replica of technical expertise: the identification of a defect in the sewer pipes in our example. To do this, companies must collect the data, in this case photographs of the pipes, and then annotate them according to the types of defects encountered. This is a first step in digitizing the activity. Then, it will be necessary to classify the data and create an ontology specific to the problems encountered by the company, with categories and subcategories of defects (corrosion with or without leakage for example).
This process of collection, annotation and classification not only provides a data set that will serve as a training set to teach the AI to recognize each type of defect, but also provides a centralized and digitized knowledge base. The latter is a valuable asset for companies, which they can reuse for other purposes.
The Transmission of Expertise
Indeed, the development of this data set does not only make it possible to train an AI capable of assisting operational staff in their technical audits. It also allows the transmission of expertise between the most senior technicians and those in training, as well as the centralization of knowledge to create a complete visual library at the disposal of all technicians.
The question of the transmission of technical expertise in field audit professions is problematic because these professions combine intellectual (making the right diagnosis) and physical (accessing confined spaces) requirements with the added repetitiveness of daily tasks. This leads to high turnover which requires continuous training of new technicians. That is why the creation of a data set that would centralize the technical knowledge specific to the business domain in a virtual library is essential. This reference database is a guarantee of the quality and durability of the knowledge acquired, transmitted by the elders for young recruits.
Once the information has been digitized and centralized, technical auditors benefit in the long term from a knowledge-sharing platform, a space for collaboration between generations and the transfer of expertise. It is also an opportunity for the company, when creating this database, to streamline the audit process. When difficult cases require the expertise of several technicians, their knowledge can be combined via this same database. The digitalization of the technical audit profession is a tremendous opportunity for collaboration between experts, transmission of skills, and assistance by artificial intelligence.
The technical audit assisted by AI offers a double benefit to companies: AI is both a support for diagnosis, which is at the heart of the business, but also the tool necessary to build a solid database, allowing to aggregate the knowledge of a company and to perpetuate the transmission of its expertise.
