How are Towercos using AI?
Yariv Geller, CEO, vHive explores how towercos are currently using AI and what the next steps might look like.
Undoubtedly, AI will change how all businesses function in the next ten years, and towercos are no exception. Yariv Geller, CEO at vHive has been working on Digital Twin solutions since he co-founded the Israel-based company in 2016. As a leader of one of the most forward-thinking technology businesses that serve the telecom tower landscape, its no surprise that he has experience helping towercos with building AI into their operations.
AI is used in a variety of activities ranging from the automation of data capture to the analysis of captured data, helping users to pinpoint items of interest in their infrastructure. “AI is something of a buzzword, frequently perceived as a magical tool” Geller confesses. “ People believe that because you are using AI, you have the ability to detect anything of interest, while in fact AI only knows what you teach it, by feeding it with significant amounts of data”. I think it’s worth breaking it down and really laying out what AI means for a towerco or an MNO.”
AI in its current form is an enabling technology that supports the automation and streamlining of entire business and operational workflows.
“One of the first steps of using AI in your operations as a towerco is for data collection,” Geller advises. AI can be used to guide autonomous vehicles – such as drones to better capture data on a tower. Better data capture means more efficient use of time and resources in the field, but moreover, the ability to capture reliable and consistent datasets, which can then be used to develop additional AI capabilities for detection of items of interest. Data that is captured manually, generates inconsistent results, that “confuses” AI, yielding poor performance.
In a digital twin context, the next step would be for this data to be transformed into a 3D visual representation of the site. This can be done with or without AI capabilities, but Geller explains that there are several new AI based reconstruction technologies that can streamline the reconstruction process by reducing the number of images required to generate the 3D model and generating better results faster.
vHive has also used telecom infrastructure specific datasets to train their solution to be able to identify the specific components and equipment on a site and build up a site inventory list. This is the first step in creating a digital twin of the site. This automated workflow, enables TowerCos to ensure accurate understanding of their sites and their associated billing – an arduous manual process otherwise.
Another item of interest that TowerCos are interested to automated detection of, is using AI for the identification of faults, anomalies and inconsistencies at a site level, that can help to identify and rectify maintenance issues in a far quicker manner. While it may take months for a field engineer to recognise a faulty rectifier, and then revisit the site with the proper tools and materials to remedy it, AI can help to immediately spot when something is broken, and generate a ticket to make sure it is repaired.
Alongside understanding the state of the tower site, one of the most compelling reasons to use a digital twin is to streamline the co-location workflow. Towercos and their customers, the MNOs can jointly work on a digital replica of the site, to coordinate addition, change or removal of equipment on the tower. Digital twins provide the freedom to visualize the site and also to manipulate it. In this case we take all the AI identified components and swap them with BIM-type components, meaning for each component a related set of attributes (such as make and model, dimensions, weight etc.) can be added. Having all this information enables to ask advanced questions in the simulated environment. AI enables to us to take the tower as it is detected in the field and to represent it in a way that it can be modified, and we can understand how this will impact the site.
“This significantly reduces the wait time for a site update for an MNO and reduces the time to revenue for a towerco” Geller says. “We’ve seen customers that would typically require a number of months for co-location request approvals, reduce this to a matter of weeks with these AI tools,” he continues.
While mainstream attention has been focused on natural language bots like ChatGPT, Geller does believe such technology can be applied by MNOs as a consumer facing technology, rather than towercos.
AI impacts the whole digitized tower flow, from ease of capturing data on site, to generate the most efficient replica of the site, the 3D Reality Model, to the ability to understand semantically all the different components the site is made of, and potential faults and then to manipulate the site and simulate different “what if” scenarios.
vHive’s system understands the context and can synchronise the data to other systems, for example integrating with the asset management system or BIM tools. To enable collaboration the data can be shared between different teams and individuals, which speeds up processes of knowledge diffusion and decision making.
Continuing to innovate in this field is a key priority for vHive, as they aim to help their customers to embrace new technology to secure revenue, find new business and streamline operations.
“We’d advise towercos to make sure they take a holistic approach to digitization and understand how different teams can benefit from AI applications” Geller stresses.
Its important for towercos to have an organised methodology and adoption of such a transformative technology needs to have strong and defined management backing as well as buy-in from different parts of the organisation.
You can meet with the vHive team at TowerXchange meetup Europe, 16-17th May 2023, and hear them discuss their solution in more depth as part of our exciting “Dragon’s Den” inspired innovation session.
* This article has been written without the use of AI
Written by Jack Haddon, TowerXchange.