South America has a new optical infrastructure player. A group of investors has launched Conecta Infra, a neutral infrastructure platform targeting high capacity intercity networks acrossthe continent. They are starting out with $350M that will go toward 6,000km of infrastructure. … [visit site to read more]

Digital infrastructure operator BW Digital announced on Friday that it has successfully landed the Hawaiki Tonga subsea cable in Vava’u, giving Tonga extra international capacity as well as needed resilience.

The 383-km cable, which officially landed on March 18, connects Tonga directly to the transpacific Hawaiki Cable linking Australia, New Zealand, Hawaii and Oregon on the US West Coast, with branching units for American Samoa and New Caledonia.

The Hawaiki Tonga cable – which has been in the works since June 2024 – was jointly funded by Australia, through the Australian Infrastructure Financing Facility for the Pacific, and the Government of New Zealand. BW Digital delivered the cable in partnership with the Government of Tonga and Tonga Cable Limited.

In a LinkedIn poston Friday, BW Digital said work is now underway to prepare the system for commercial service in the coming months.

The Hawaiki Tonga cable is Tonga’s second international subsea connection, and its first direct link to a major transpacific cable system. Up to now, its only source of international subsea capacity has been the Tonga Cable – owned by Digicel Tonga, the Government of Tonga and Tonga Communications Corporation – that connects to Fiji.

The lack of redundancy has made Tonga highly vulnerable to internet disruptions. In January 2022, the eruption of the Hunga Tonga-Hunga Ha’apai volcano damaged both the Tonga Cable and the Tonga Domestic Cable Extension (TDCE), the island nation’s sole domestic subsea cable that connects the main island of Tongatapu with the northernmost island of Vava’u.

The TDCE suffered damage again in 2024 due to an earthquake.

Partner Article

Huawei unveils AI-native framework and new generation solutions to enable all intelligence 

The communications industry is entering a new phase of transformation, as operators move beyond basic automation towards building intelligent networks. Rising network complexity, surging data traffic and the rapid adoption of Artificial Intelligence (AI) across enterprises and consumer use cases are exposing the limitations of traditional operations models, which remain largely reactive and fragmented.  

At the recently concluded MWC Barcelona 2026, Huawei outlined its vision for this shift, unveiling what it describes as the industry’s first AI-native framework for intelligent operations, alongside a new suite of solutions designed to accelerate the transition. The announcement is in line with a broader industry trend, where telecom operators are increasingly looking to leverage AI across the entire network lifecycle, from Operations and Maintenance (O&M) to customer engagement and monetisation. 

According to Capgemini, telcos have, on average, achieved a 20% improvement in operational efficiency and 18% reduction in operational expense through autonomous networks over a period of two years. While telcos are increasingly using AI to automate networks, the rise of AI-powered tools and applications is placing new demands on network performance and reliability.  

President of Huawei’s Global Technical Service, Bruce Xun, announced the launch of three solutions, Agentic BSS, SmartCare Intelligence and AUTINOps, which he believes will “ensure a seamless evolution while establishing new paradigms, creating new value, and achieving a massive leap in intelligence.” He was speaking in a session on Agentic Services and Software Enable All Intelligence.  

These solutions are based on an AI-native framework, which is “designed to accelerate the transition of AI innovations into real production, creating tangible new value,” said Xun.  

This framework is built on three core pillars. To begin with, it targets operational and business challenges that legacy solutions have failed to resolve. Secondly, it leverages digital twins and telecom-domain models to enable optimal solutions. Thirdly, the framework reinvents workflows and talent development to ensure a seamless collaboration between human experts and digital employees. 

A key challenge for service providers today is that traditional operations models remain largely reactive.  The tariff design and launch process is slow, and may not meet customer needs or effectively recommend the right products at the right time. 

In this context, Agentic BSS focuses on transforming business operations by introducing AI agents that can interpret customer intent, design new offerings and optimise customer engagement. By enabling collaboration between multiple agents, service providers can significantly reduce time-to-market and deliver more personalised services. 

“The key to intelligent operations lies in our ability to anticipate customer intent and identify unmet needs,” said Xun. This shift towards intent-driven operations is particularly relevant as operators expand into enterprise services and digital ecosystems, where responsiveness and customisation are critical.  

Huawei received a Silver award in the Total Experience category for its collaboration with a Chinese operator, where AI-driven account management agents were introduced. These digital assistants improved response times, reduced business processing time by 30%, and enhanced enterprise customer engagement. 

Reimagining network optimization  

While business transformation is one aspect, network operations remain central to the AI-native vision. SmartCare Intelligence, another solution introduced by Huawei, is designed to move network optimisation from a reactive to a predictive model. By leveraging large AI models, including User Experience Large Model (UELM) and Beam Space Large Model (BSLM), the system can analyse network data in real time, simulate potential scenarios and generate optimal adjustments. This approach ensures that the networks are always tuned to deliver optimal performance. This is significant because service providers are struggling with growing network complexity. AI-driven systems can help them provide consistent service quality.  

Towards predictive and preventive operations 

The third pillar of Huawei’s portfolio, AUTINOps, focuses on Operations and Maintenance (O&M), introducing a predictive approach to network reliability. 

Traditional O&M models often respond to faults after they occur, leading to service disruptions and longer recovery times. AUTINOps addresses this by identifying potential risks in advance and taking preventive action. It uses a cross-domain digital twin network and EDNS 2.0 model to monitor networks in real time, detect risks and trigger recovery actions. It combines proactive risk mitigation (T-1) with quick fault recovery (T0) to ensure “dual protection,” leading to high uptime and network reliability.  

Huawei’s approach is beginning to see validation in real-world deployments. At the World Communication Awards (WCA), the company received a Silver award for Best Digital Transformation Programme for its work with a Middle Eastern operator, where it implemented an AI-driven O&M system focusing on fault management and resource topology optimization. Based on the AUTIN platform, the solution enabled over 95% topology visibility across VoWiFi access points, LTE anchor points and IMS core network elements. It delivered measurable improvements, including service availability exceeding 99.5%, a 40% reduction in call drop rates, and significant gains in fault detection and repair times. 

Xun emphasized the importance of robust connectivity, without which AI remains an information silo. He urged the industry leaders to standardize definitions and specifications of AI native elements, jointly design new business and transaction models and to share best practices to simplify network operations while ensuring improved performance.  

Vodacom Lesotho is investing more than US$40 million (LSL 700 million) to upgrade its network infrastructure and expand connectivity, as it looks to improve service quality and extend coverage to underserved areas.

According to TechAfrica News, the programme will focus on boosting network performance and widening access, in a move aimed at strengthening digital inclusion across the country.

CEO Mohale Ralebitso said the operator is also introducing artificial intelligence across its operations, spanning network management, customer engagement and internal processes. The use of AI is expected to drive efficiency and support more data-led decision-making.

The investment is set to deliver broader coverage and improved service quality for customers, while also enhancing operational performance through advanced analytics.

Alongside its network push, Vodacom Lesotho is expanding its financial services arm, VCL Financial Services, positioning it as a more independent unit offering savings and investment products. The move is intended to deepen financial inclusion and provide customers with more flexible financial tools.

Globe Telecom has expanded its LTE and 5G footprint in the Philippine province of Tarlac, as it moves to meet rising demand for faster and more reliable connectivity.

According to the Manila Standard, the upgrade covers 18 municipalities, including Tarlac City, Capas, Concepcion, Gerona, Victoria, Paniqui, Mayantoc, San Jose, Moncada, Santa Ignacia, La Paz, Camiling, Bamban, San Clemente, Pura, San Manuel, Ramos and Anao.

Globe said the expanded network will enable improved browsing speeds, clearer voice calls and faster data services for residents across the province.

Joel Agustin, Globe’s head of service planning and engineering, said the rollout supports the region’s mix of economic activity and tourism, adding that the operator is focused on delivering more seamless connectivity for users at work, in education and on the move.

This Industry Viewpoint was authored by Stephen Lafaille, VP Business Development

The data center industry is entering a defining era – one in which power, not land or capital, determines competitive advantage. As AI compute requirements surge and the grid strains under unprecedented demand, operators face a simple but unforgiving equation: every megawatt not available for IT constrains … [visit site to read more]

Huawei has urged telecom operators to embed artificial intelligence across networks and services as the industry moves beyond the mobile internet era into what it describes as the “Internet of Agents”.

Speaking at Mobile World Congress Barcelona, Richard Liu, Huawei President of ICT Marketing & Solution Sales, said AI will fundamentally reshape telecom networks, services and operations, creating new opportunities for operators to generate revenue and improve efficiency.

While telecom networks have expanded rapidly over the past decade, operators now face growing challenges including increasingly complex architectures, high operational costs and limited differentiation between services.

As AI evolves rapidly, we are shifting from the mobile internet to the Internet of Agents, and AI will increasingly be embedded across devices and applications, with trillions of intelligent agents expected to emerge.

Telecom operators are well positioned to take advantage of this shift because communications networks are already highly digitalised. It’s suggested that carriers could both apply AI internally to improve efficiency and provide AI-enabled infrastructure to support digital transformation across industries.

Huawei is positioning AI as a core component across three layers of telecom networks: services, network operations and network infrastructure.

At the service level, AI is moving from being optional to a necessity. Operators are increasingly integrating AI capabilities into core offerings such as voice, mobile data and home broadband services in order to create what Huawei calls “AI-native services”.

Huawei is developing multi-agent collaboration platforms designed to allow operators to deploy AI-powered agents for areas such as call handling, customer experience management and home broadband networks.

At the network level, AI is also being integrated into telecom infrastructure to enable greater automation. AI-driven networks are evolving from automation in specific operational and maintenance scenarios towards more advanced forms of autonomous networking that can optimise performance and manage resources across multiple domains.

These capabilities could help operators reduce operational costs while improving energy efficiency and enabling more personalised service experiences for users.

A key theme of Huawei’s strategy is “experience monetisation”, which focuses on charging for guaranteed network performance and service quality rather than simply providing connectivity.

Technologies such as 5G-Advanced, high-capacity fibre access and fibre-to-the-room deployments are enabling operators to deliver differentiated services to both mobile and fixed broadband customers.

For consumer markets, operators could offer guaranteed performance for high-value users in scenarios such as dense urban environments or high-speed mobility. In the home broadband segment, AI can be used to optimise network performance and provide automated troubleshooting capabilities.

Huawei also highlighted opportunities to enhance traditional telecom services with AI. AI-powered voice services could improve call clarity and introduce features such as real-time translation, automated meeting summaries and intelligent call assistants.

Similarly, AI-enabled broadband services could allow users to diagnose and repair home network problems through simple voice commands, while also enabling personalised content recommendations and enhanced video communication services.

Beyond consumer services, Huawei sees strong growth potential in the enterprise market, particularly among small and medium-sized businesses.

Operators could combine connectivity with computing, storage and AI capabilities to deliver integrated digital solutions for businesses such as retail stores, manufacturing firms and schools. These solutions could include applications such as video security systems, smart retail tools and AI-assisted education platforms.

Huawei also highlighted its ongoing work on autonomous networks, an industry initiative aimed at increasing network automation. The company said its technologies are already deployed on more than 130 networks worldwide to automate tasks such as fault management, network optimisation and energy efficiency improvements.

Huawei is working with industry organisations including the TM Forum to advance standards for autonomous networks as the industry works towards Level 4 autonomy, where networks can operate with minimal human intervention.

Additionally, Huawei continues to invest in AI-driven algorithms and infrastructure technologies across areas such as radio access networks, optical transport and AI computing platforms in order to support the next generation of intelligent telecom networks.

Interview

We spoke with Liu Chao, CEO of the Huawei’s Manufacturing & Key Enterprise Account business unit, about the seismic impact AI is having for the manufacturing industry

At Mobile World Congress 2026, AI finally appeared to be coming of age. From myriads of commercial AI agents to early demonstrations of physical AI, it was clear that AI was finally becoming

For Huawei’s Liu Chao, the era of treating AI as a high-tech accessory is over for the manufacturing sector.

“AI is now more than tools,” said Liu in an interview with Total Telecom. “It can be a unique distinguisher for manufacturers to set themselves apart from their competitors[…] AI is now becoming an important paradigm shift in innovation and in leadership.”

““This shift is being driven not only by the growing maturity of AI, but also by the urgent need for manufacturers to strengthen their competitiveness. Established leaders in traditional manufacturing sectors, such as automotive, are facing increasing pressure as more players actively embrace advance technologies.

Given the precision and high standards required in manufacturing, industrial players place a strong emphasis on proven reliability and predictable outcomes. This means they tend to wait for new technologies have demonstrated clear value and stability. For Liu, this urge to wait is a “trap”.

“Adoption of AI is not optional. It’s a mandatory choice you have to make. The question is not whether to do it or not, but how to do it,” he said.

Bridging the expertise gap

Perhaps the biggest hurdle to adoption, Liu explained, is the lack of cross-discipline expertise. Industrial experts are typically not AI experts, and vice versa,

“I think one of the key priorities for manufacturers adopting AI is deepening their understanding of the technology and its evolving trends,” said Liu. “This also means strengthening capabilities in data and digital infrastructure, while developing more talent with AI and IT expertise – both of which are essential to fully unlock the value of AI.”

For AI adoption to scale across industry, both manufacturers and tech companies need to cultivate multidisciplinary talent that combines both industrial and digital expertise.

“We need AI experts who have the knowledge and background in the manufacturing sector,” he said.

It is only with this shared expertise, Liu argues, that the industry will be able to develop AI models tailored to the manufacturing sector’s specific needs.

“General models like OpenAI answer questions based on public information. But when it comes to the data about a specific company, industry, or process, these models are not good at giving very specific answers,” said Liu. “In manufacturing companies, the data about operation management, production processes, and research and development is proprietary and private. So, they need specialised solutions.”

Practical first steps: Pilot projects and infrastructure foundations

With this in mind, what does early AI adoption look like for manufacturing companies?

For Liu, initial focus should be not on overall transformation, but on addressing specific challenges.

“When a manufacturing company comes to us and says they want to begin using AI, we first discuss their pain points in their business,” Liu said, noting that identifying the right use cases can generate early value.

“We have to find some typical cases where AI can be applied and give a quick win to our customers,” Liu said. These early projects often act as pilot programmes that help organisations build internal experience and refine their data strategies.

“In the first stage we identify scenarios as the first batch of AI adoption pilots,” Liu explained. “Then in the next step we review their more confidential or private data in production or R&D and help them standardise it, ready for use in AI models.”

Automotive taking a lead

One manufacturing industry leading the pack when it comes to AI adoption is the automotive industry.

“Each year in China, 50% of new cars are connected to the internet and are electric vehicles. The changes in the market are very fast. These days, auto manufacturers are launching their new car models almost as frequently as mobile phone makers are launching phones,” he said, adding that “autonomous driving and smart cockpit capabilities are all enabled by AI models.”

The most advanced carmakers are using AI across product development, factory operations, and quality inspection. This is allowing customers to enjoy a far greater level of personalisation as part of a C-to-M (Consumer-to-Manufacturer) framework.

“It is an end-to-end process that allows full customisation by the consumers,” explains Liu. “It’s how auto manufacturers in China are trying to win in such fierce competition.”

“In the assembly line, a fully assembled car is built every minute,” Liu continued. “When the customer chooses a specific configuration – say, for example, a yellow safety belt – you have to make sure that yellow belt arrives at exactly the right point in the assembly process. That needs AI-enabled scheduling with the data flowing from the order side directly to the production.”

Networks come first

Of course, a strong foundation of digital infrastructure is a critical requirement in this journey.

“The precondition is that you have very solid network connections and very good hardware,” Liu said. “Without this, putting AI into action is incredibly difficult.”

For Liu, the pace of change means manufacturers must continue learning and adapting as AI technologies evolve.

“You cannot wait for the latest technology for fear of being left behind because AI is changing so quickly,” he said. “You have to learn throughout the process of adoption.”

Keep up to date with all the latest telecoms news with the Total Telecom newsletter

Also in the news
World Communication Award Winners 2025
Ofcom clears the way for satellite-to-smartphone services
LG Uplus’s AI voice call app glitch leaks user data