understanding of what is happening within the network .
In terms of general tech support though , at Bridge we have founded our whole approach on business as a person-focused practice ; we build relationships with clients , collaborators and even competitors not only for the purpose of selling products , but in order to collectively
“ Broadcasters need to consider their operational context before implementing Cloudbased services .” – Bridge Technologies
contribute to pushing the industry forward . As such , we can say with some certainty that when you contact the Bridge offices for support , it will be a real person you speak to . Interra Systems : Over the past year , the media industry has seen a marked rise in automation across production , output , and localisation services , with AI and machine learning playing an increasingly pivotal role . As workflows evolve , there ’ s a greater understanding of where these technologies can be most effectively applied .
In particular , AI plays a transformative role in T & M within the media industry by streamlining and enhancing critical processes like captioning and subtitling . As broadcasters and global streaming platforms strive to overcome language barriers , AI and ML have become essential tools for ensuring the accuracy and accessibility of subtitles and captions . By automating these processes , AI not only improves efficiency and precision but also elevates the overall user experience , making foreign-language content more accessible and enjoyable for global audiences .
With this surge in automation , the importance of monitoring and maintaining these processes has grown . To manage intricate workflows and multiple formats , ensuring fast , reliable diagnostics , analytics , and quality checks at every stage of production is essential for smooth operations . Torque Video Systems : While generally , I think the jury is still out on where ML and AI fit into T & M for broadcast , there have been numerous advances in academia on the application of AI in these areas . Overall the rapid advance of AI into everyday life over the past 12-18 months has been nothing less than astounding .
From personal experience , 2024 marks the year that I no longer first reach for a search engine when looking for answers to complex questions . In the past , it is very time consuming searching the web for an elusive answer to difficult or complex questions . Now , almost exclusively , I will ask one or two
AI chat bots first and review the responses . Usually it is the correct answer . ( Well , usually , but asking the same question to multiple AIs allows me to find the right answer quickly ).
The telecom industry has been working on AI-based network monitoring tools and help desk assistants for years . In those cases , the network monitoring tools can make a difference given the massive number of devices , network points and transmission links . However , broadcast networks are , in order of magnitude , more simple . So , will AI make sense there ? I ’ m not so sure .
One area of research I have seen gaining traction is the use of AI models to evaluate picture quality of still or moving images . Neural networks excel at identifying and matching patterns against a pre-trained ‘ source of truth ’. Theoretically , that is a perfect fit for looking at video and identifying quality issues .
The next challenge for T & M and monitoring systems would then be to tie the identification of a quality issue with the corresponding root cause within the network . I see those kind of semi-automated systems coming soon .
One other possibility is for the analysis of large quantities of end user data to distil viewing trends and preferences . Those kind of big data ( and Big Brother ?) insights can , in turn , drive recommendation engines as well as which ads are displayed to which viewers .
As to whether GenAI has a role , such as in tech support , sadly , yes , and it is already happening . Just try to reach a ‘ real person ’ at your bank , insurance company , or an airline these days . More often than not , you are stuck with a chat bot that is really not useful .
Beyond tech support , there are a few good causes where GenAI could be applied . For example , it could make customer service interactions faster by pre-filling forms and providing more user-centric and targeted product promotions . VeEX : AI ’ s superpower is pattern recognition , which may allow it to identify issues before they become problems . However , it requires that us humans analyse , interpret and document a large number of cases , to get the training started . The capabilities are already there , but the industry in general has not been good at documenting and collecting the data or providing the right context for each individual case . Centralised monitoring systems may be the first ones to take advantage of AI . Large amounts of test records do exist , especially from CSPs that have been using monitoring systems and workflow processes over many years . However , ( a ) they are considered confidential intellectual property and ( b ) they may have a pass / fail resolution but may lack the background information to provide the necessary context ( unlike other fields , like medical , do ). Simpler , very focused solutions are already being offered . Without large open datasets to train from , the self-learning solutions will develop at a slower pace , as they generate their own databases .
T & M tools that are analogue in nature require a large database of varied , real trace signatures to learn from as simulated signatures are inadequate to teach pattern recognition . This requires significant costly embedded processing power which opposes the efforts of Service Providers to reduce capital expenditures spent on tools .
AI can be used for Tier1 tech support prescreening by going through a list of common operational mistakes that might impact service activation or performance . Witbe : AI isn ’ t a magic solution , but it has already delivered tangible benefits in testing and monitoring through enhancing human capabilities – not replacing them . AI and machine learning have been instrumental in creating more human-like automated testing scenarios , helping systems better mimic
real user behaviour and decision-making processes . In practical terms , AI has already allowed test scenarios to be scripted up to ten times faster . It has also enabled testing at a scale impossible for human teams working by themselves , which is particularly valuable given the number of different devices and operating systems that services are expected to currently support . The key to success lies in thoughtfully integrating AI into existing workflows , where it excels at streamlining repetitive tasks and pattern recognition , and leaving complex decision-making to human experts .
Will we see a new era of ‘ self-diagnosis ’ and ‘ self-healing ’?: Aprecomm : With the advent of AI and machine learning , telecom networks can care for themselves . They self-optimise and self-heal to give broadband subscribers the best online experience possible , wherever and whenever , across all their devices and applications . Today , artificial intelligence and machine learning systems can identify and resolve many issues in the WiFi and core access network , whether by making channel changes on the fly to mitigate interference
EUROMEDIA 11