Neural network
We had to teach our neural network to identify the following situations:
- Incoming customers are not included in the accounting system (face recognition to control entry and control the number of visitors.
- Services rendered do not coincide with those services that are included in the accounting system. It was for this task that it was necessary to recognise people in chairs and control the dentist’s working time.
- Non-compliance with medical personal protective standards (the use of shoe covers, gloves etc).
- Violation of the work schedule (time and frequency of absence of the employees from the workplace / front desk);
CVC implementation process summarised
To obtain accurate results, the program had to learn to recognise a whole set of various scenarios:
"a person in a chair is covered / or not";
"the presence of a doctor near the chair",
"being near the chair of an assistant";
"the position of the doctor’s hands in the area of the person’s oral cavity in the chair".
instant results
In the first week of delivery of the tool, the following situations were recorded in the clinic
- A dentist personally received payment from the client (although only the receptionist had the right to do this). As a result, the payment was not recorded;
- A treatment service was provided to the client, and only an examination fee was applied to the bill
- The x-ray machine was used multiple times, yet there was an absence of items on the billing system;
- The multiple absence of the receptionist at the front desk for more than 10 minutes.
There were no instances recorded of health and safety violations.
