Cover Story allow the detection of complex issues thanks to its capacity to handle large sets of data and correlate these metrics to alert on both abrupt and slow degradations in a blink of an eye . Moreover , AI will enhance the connected experience by providing personalised improvement actions tailored to the customers and its devices , thanks to massive learning from the entire installed base of devices .” MAINTENANCE . “ AI can facilitate predictive maintenance , detecting and anticipating devices that will likely require repair ,” says Ferré . “ For example , a washing machine that is becoming noisy and thus needs repair or a fridge that starts to consume more energy and is likely to require a fix . AI can also help define when an appliance in a specific room becomes too noisy for those users . One easy and obvious use case that we have already implemented at Vantiva is a gateway and set-top box that can analyse its own functioning and anticipate a problem .”
Cover Story allow the detection of complex issues thanks to its capacity to handle large sets of data and correlate these metrics to alert on both abrupt and slow degradations in a blink of an eye . Moreover , AI will enhance the connected experience by providing personalised improvement actions tailored to the customers and its devices , thanks to massive learning from the entire installed base of devices .” MAINTENANCE . “ AI can facilitate predictive maintenance , detecting and anticipating devices that will likely require repair ,” says Ferré . “ For example , a washing machine that is becoming noisy and thus needs repair or a fridge that starts to consume more energy and is likely to require a fix . AI can also help define when an appliance in a specific room becomes too noisy for those users . One easy and obvious use case that we have already implemented at Vantiva is a gateway and set-top box that can analyse its own functioning and anticipate a problem .”
As to the role AI could play in maintaining a secure home network and protecting the household from malicious or inappropriate content and whether there are any customer data risks , Walls notes that many network and application security vendors are already doing this . “ The key is ensuring that the applications have the data they need to protect each subscriber and distribute threat mitigation and information when made aware of threats that have been seen elsewhere . In today ’ s home user environment often the security threats are from third party consumer , smart home and IoT devices . These are frequently not a managed service from the service provider .”
“ From a more general security perspective , the ability to be able to learn malicious actors ’ patterns and behaviours against norms is constantly being improved at both the Cloud and closer to the first packet ingest at the CPE level ,” advises Cheevers . “ With the ML models driving new local templates pushed to the CPE device to identify the security issues at the GW before getting deeper into the network , this has become the norm of current CPE security .” PRIVACY . “ From a Privacy perspective , there is huge opportunity to improve consumer privacy from rogue listeners and camera devices and to make consumers more aware as to where audio or camera devices may be transmitting outside their home network . There has been much discussion about trying to create local file / data vaults for consumers that remain local to their home for sensitive
“ CPE has all the necessary hardware capabilities to host household AI .” - Lionel Gremeau , SoftAtHome information vs relying fully on Cloud storage solutions today ( photographs , scans of personal documents , etc .). But these solutions also come with issues , such as lack of redundancy in the case of home fire / flood and a lack of access remotely ,” he warns .
“ AI plays a pivotal role in cybersecurity and content filtering ,” asserts Hendricks . “ It excels in identifying devices with MAC randomisation , ensuring compliance with security and internet access control policies . AI can also detect anomalies in devices , particularly IoT devices , and quarantine them accordingly . “ Safeguarding customer data is of utmost importance to Gamgee . Robust data protection measures , strict anonymisation practices , informed consent protocols , and compliance with privacy regulations ensure that customer data remains secure within the AI-enhanced CPE ecosystem .”
“ AI will help households move beyond static rules ,” suggests Cottam . “ The ability to vary rules intelligently , to proactively protect the home without a change in rules being necessary , and to provide easier control of home security through voice commands ( as opposed
to complex menus ) will all add value and make the home safer . A growing area of interest is the scraping of customer data , which is not necessarily malicious but certainly highly intrusive . It can also lead to embarrassing or risky data leakage . Home AI could protect against this by preventing silent or thoughtless leaking of data from the home , and restricting access of external AIs to home data , putting householders back in control .” RULES . “ Unlike the rules-based approach typically utilised today , AI uses machine learning algorithms to advance threat detection and content filtering ,” advises Bensaid . “ By learning user behaviour and establishing baseline
“ AI can surpass traditional approaches to service quality and provide a more seamless and satisfying home experience .” – Arnaud Bensaid , prpl Foundation
patterns , it can analyse network traffic , identify any anomalies , and proactively detect threats and breaches in real-time . Through content evaluation based on pre-defined criteria , AI can also filter out harmful and inappropriate material and deliver a safer online experience . The standardisation of the CPE software stack by prpl Foundation plays a crucial role in amplifying AI capabilities to create a safer digital environment for households .”
“ Cyber-attacks have been rising since the pandemic ,” reports SoftAtHome ’ s Gremau . “ CPEs represent the operators ’ first line of defence and are subject to numerous and multifaceted attacks . We now need more than advanced firewall rules . AI can play a key role as it can learn from normal user usage patterns and detect unexpected traffic : to or from an unexpected destination , on unusual ports , with an abnormal amount of traffic ( cannon attack with max bandwidth upstreaming from cameras , for example …). In that case , there is no specific customer data as it is learning and monitoring usages without extracting private customer information . On the other hand , all data manipulated locally on the box are not sent to the operators ’ backend and don ’ t break any GDPR rules or customer privacy .”
“ AI can provide enhanced cyber protection by detecting an unusual event ,” says Ferré . A holistic approach to security is to generate an alert each time something “ different ” happens . Moreover , generative AI can lure cybercriminals and catch them in the act of trying to breach a system . AI can be much more subtle when it comes to detecting threats or risks , with a much lower need for expensive and cumbersome hardware . For example , a camera is of no real use for security . Intrusions can be detected with sound and electromagnetic fields . Once an intrusion occurs , a camera is ineffective .” SERVICES . As to what additional services or functions could a more intelligent CPE provide in future , Walls suggests the sky is the limit . “ The app-enabled services gateway project from Broadband Forum helps service providers effectively deploy , activate and manage third party applications such as parental controls , Wi-Fi analytics , streaming services and security solutions to consumers on a centralised Internet gateway device or other CPE ,” he advises . “ AI in partnership with an ASG
14 EUROMEDIA