The research indicates that 94% of the service providers surveyed plan to use software-defined networking (SDN) technologies within their automation strategies.
Furthermore, 92% consider the integration of artificial intelligence and machine learning (AI/ML) important or critical for developing these capabilities.

Nevertheless, the results show technical and operational difficulties that condition the deployment of systems capable of analyzing the state of the network and executing actions in an automated way.

83% Identify Data Problems.
One of the main obstacles highlighted in the report is the availability of consistent information from different network systems.
83% of the service providers surveyed consider poor or inconsistent data a significant barrier to implementing AI in their infrastructures.
Broadband networks typically combine equipment, software platforms, cloud services, and management systems from various vendors.
To use this information in automated processes, systems need data that can be interpreted consistently and related to network resources, services, and events.
The report links this need to the use of common data definitions, identifiers, and models that allow relationships to be established between the different components of the infrastructure.

SDN as a Foundation for Automation:
Research places SDN technologies among the elements operators are planning to use to develop networks with greater levels of autonomy.
94% of participants state they plan to use SDN within their strategy.
This approach allows for the separation of certain control functions from the physical infrastructure and the management of network behavior through software.
AI and machine learning can then be used to analyze operational information and support decisions regarding service operation.
However, the adoption of SDN and the incorporation of AI do not, in themselves, guarantee that a network can operate completely autonomously. The degree of automation also depends on the available interfaces, data quality, and the systems' ability to execute the planned actions.
Predictive Maintenance and Anomaly Detection:
The report identifies several functions where operators are already using or plan to use AI and machine learning.

Operational applications of AI/ML:
Predictive maintenance: 90%;
Anomaly detection: 72%;
Automatic fault resolution: 69%

Percentage of providers that use or plan to use AI/ML in each function, according to the cited report.
Predictive maintenance, mentioned by 90% of participants, appears as the most widespread or planned operational application.
This is followed by anomaly detection, at 72%, and automatic fault resolution, at 69%.
These figures include both current and planned uses, so they do not exclusively represent deployments already in operation.
Customer Experience and Service Assurance
: When asked about the main use cases for AI, the report identifies customer experience optimization, mentioned by 87%, and service assurance, cited by 85%.
These areas relate to monitoring service quality, analyzing incidents, and managing conditions that affect users.

78% cite interoperability problems.
Compatibility between devices and systems from different vendors is another obstacle to automation.
78% of respondents identify the lack of interoperability between manufacturers as a significant barrier.
The problem relates to the need to integrate control and management platforms that use different interfaces, data models, and operating procedures.
This heterogeneity makes it difficult to apply common automation mechanisms to all elements of a network.

Integration with existing DSL, HFC, and fiber networks:
90% of the providers surveyed consider integration with existing systems to be a significant challenge.
The report specifically mentions DSL infrastructure, cable/HFC networks, and different generations of fiber optic equipment.

Operators must integrate new SDN and automation platforms with these technologies without assuming that all equipment has the same control capabilities or management interfaces.
The coexistence of multiple technology generations adds compatibility requirements that must be addressed during implementation.

Standardization and Interoperability Testing:
The report highlights the role of standards in establishing common interfaces, definitions, and models among systems from different manufacturers.
The Broadband Forum links this standardization to the possibility of integrating hardware, software, and management components within automated architectures.
The text also points out the importance of complementing specifications with reference implementations, interoperability testing, and proof-of-concept projects.
These activities allow for verifying system behavior in specific scenarios and detecting integration problems before scaling up deployment.
The research does not identify in the provided excerpt which specific standards were evaluated, nor does it offer interoperability test results.

Limitations of the published data
The percentages in the report show a favorable orientation towards SDN and AI/ML, but also identify barriers to their implementation.