France Is on Its Way to a High-Quality Fibre Optic Network
#France Is on Its Way to a High-Quality Fibre Optic Network Network
PUBLISHED ON 9/9/2023
France has made significant progress in the deployment of its fibre optic network over the last 10 years and has taken a European leadership position in switching to full fibre technology. In total, more than 4 out of 5 households in France can now enjoy the luxury of a fibre optic connection, and the number of active subscriptions continues to grow at a steady pace. France is gradually approaching the strategic goal of switching to full fibre by 2030, which is a deadline for nearly 15 million French people currently connecting to the internet via ADSL, to switch to fibre technology.
A fly in the ointment
However, the quality of deployed networks, as well as the quality of daily operations remain an active topic for public discussions. First, as reported in April by Avicca, 28 local authorities representing 12 million inhabitants, co-signed a joint press release in the form of a warning to major telecom operators and their subcontractors. They requested immediate improvements in network quality to prevent equipment deterioration.
In July, Senator Patrick Chaize seconded them and tabled a bill aimed at strengthening controls and sanctions against telecom operators responsible for last-mile connections. The Senator’s focus was on the quality of the Public Initiative Networks, stating that at least 80% of the connection points would be “more or less seriously damaged.”
As a result, in December, ARCEP, the Regulatory Authority of France for Electronic Communications thoroughly examined the fibre optic network equipment in 50 departments of France. As their report stated, a "significant proportion" of faults in fibre optic equipment could lead to quality of service problems in the medium term.
Based on the public comments of French infrastructure operators, 70% of the connection failures are due to technicians, and only 30% are due to the network. This proves that the French industry knows the points for improvements and can come up with a remedy plan. The plan, taken as guideline while deploying and exploiting networks, was based on utilising the procedure of visual control over the infrastructure units all over the country, i.e., continuously taking pictures to prove the state of the infrastructure unit before and after each operation. However, these came in huge amounts, summing up to 10 million operations a year, or 70 million pictures to be processed – the task a human eye could hardly cope with.
As the next step in fighting the manpower quality problems, technicians’ work results, and the work of the increasing number of subcontractors, local ISPs and infrastructure companies employed the image analysing tool based on artificial intelligence algorithms. The tool is capable of analysing thousands of images per minute and detecting deviations and anomalies with 95-98% accuracy.
These images of wiring cabinets, splice boxes and optical sockets are sent by technicians while being on-site. Few pictures are taken before the operation to detect the initial state of the equipment, and few after the operation. All pictures are then uploaded to the AI-driven platform, which carries out immediate real-time analysis and returns the verdict on the points of improvement while the technician is still on site.
Inveniam's AI-driven solution by Inveniam took a leading role in this process, helping infrastructure companies like Axione and the local ISPs to improve the quality of operations. Within the first year of using the Eygles software, some customers found out that up to 70% of on-field operations are performed with some error or anomaly.
This means that because of these errors rework costs inevitably increase. And the additional rework costs also bring down company sustainability targets. However, with the help of Eygles software, the number of internal quality issue tickets went down by more than 45% by the end of the year.
Do you have issues related to quality control in your telecom infrastructure?
Do you have questions about the deployment of artificial intelligence in your organization?