How digital twins are transforming the way towercos are accumulating and managing data
2021 F-1 has been controversial, to say the least!
Did you get a chance to witness the last lap of the final race of the season? Verstappen won, Hamilton didn’t. Yes, drivers take up all the media attention. But, at the back-end, there is a colony of men, women, machines, and digital twins at work.
The digital versions of Formula One race circuits and cars are used to win races by simulating various versions of the race even before they take place. Once multiple variations are researched and simulations carried out digitally, a perfect winning strategy is drafted and followed to a T to win the race.
At first glance, it could seem something straight out of science fiction but digital twins are very much a reality now. In fact, about 50 years ago, NASA made its first attempt at creating a digital twin, even though the term caught on later, to assess and replicate the conditions aboard Apollo 13 after damaged wire insulation had caused its oxygen tank to explode mid-mission. The digital replica on-ground made it feasible for the engineers to test distinct rescue solutions to get the astronauts back to earth safely.
With growing interest and advancement in the digital twin technology, these digital versions of physical assets and scenarios or “make-believe realities created by a bunch of computers” as Gene Kranz, Chief Flight Director for Apollo 13, called them, hold the potential to profoundly impact our lives and the way we do business.
Digital twins for towercos
When it comes to digital infrastructure, accessing and tracking the performance of billions of dollars in managed assets of the tower industry, spanning across geographies and thousands of sites can seem impractical. Manual inspections of these assets and legacy systems of data storage can only take you so far. What towercos need is a comprehensive solution that not only offers them a formula for easy data capture and storage but also lets them centrally monitor, compare, and even predict the changes in their tower assets. Digital twins that take into account the physical attributes of these tower assets to the environmental conditions they are placed in are a time and cost-saving option that make this achievable and take towercos another step closer towards realizing absolute digital transformation.
Simply put, this digital twin solution comprises of two parts- the physical and digital versions of the assets and the data flow between them.
The tower infrastructure and assets are mapped and the data captured is modeled into a digital replica. Once the data from the physical asset has been used to create the digital one, data from the digital twin can then be used in the physical reality to improve business processes and to make the transition from solely relying on taking preventive maintenance measures to predictive ones.
Needless to say, the success of this digital twin solution hinges on data making the accumulation and digitalization of all tower data the need of the hour.
Data that towercos can map to generate digital twins
By now, we understand the importance of data for successful digital twin modeling, and before towercos and MNOs can dive into capturing it, a proper data-accumulation strategy needs to be laid out that defines the data that will be required for short-term analysis, decision-making, and long-term data transformation so that inspection activities can be timed properly and any data noise can be avoided.
For the creation of digital twins, asset data sets need to be defined to capture everything from the installation date of the asset, collection date of data to the asset’s physical characteristics such as dimensions, the orientation of load, surface color and temperature, and environmental conditions within which the asset has been placed. Such a virtual representation of towers and assets not only helps towercos identify potential spaces where additional antennas could be added for improved efficiency, but it also helps to identify if the tower is overloaded placing the structural integrity of the tower at risk.
Once this captured data is fed to an AI-powered digital twin model that is programmed to create custom inspection algorithms, towercos will be able to predict dimensional changes in the steel lattice, environmental stress due to heat, wind load on the tower infrastructure, and the onset of cracks and rust. This would take predictive maintenance to a higher level of efficiency, significantly reduce the downtime and maintenance costs, and increase the overall customer experience that towercos can offer to the MNOs.
Telecom site infrastructure mapping for the creation of digital twins requires high-resolution pictures and panoramic videos of everything from the site compound to the assets hosted within the compound from a photogrammetric point of view i.e., of predefined points on the infrastructure from specific angles. These can then be overlapped to create realistic 3D models while keeping the amount of data being captured at its optimum, thus also cutting down the amount of time it takes to capture this data as well as to review it.
Detailed mapping also provides an automatic inventory and detailed attributes of all the towerco-owned and MNO-owned assets which can then be tagged to the owner company for maintenance and automation.
After the data capture and modeling of the digital twin is done, being able to convert that into actionable intelligence is where the true value of a digital twin lies. The tower owners and MNOs need to be able to leverage this data to stay current on the changes in their infrastructure and assets, preventive and predictive maintenance requirements, the possibility of asset failures and respective downtime, energy consumption and costs, and the potential for business growth and diversification.
Even though towercos are adopting automation of various assets such as diesel gensets, heating, ventilation, air conditioning, etc. to control energy costs, the digitalization of the mapped data is still limited to the towers and the loads placed on them. This could be because, as opposed to asset mapping, infrastructure mapping provides quick wins such as reduced opex.
Asset mapping can, however, in the long-term help with the long-tail analysis of identifying assets and sites that are or could be problematic and assist in having precautionary maintenance measures in place to avoid any downtime. This also helps in keeping track of the condition of assets to determine which ones can be sweated and which ones need to be serviced to increase their lifetime value. Keeping tabs on diesel consumption and feeding the data to a digital twin could help identify genset inefficiency or fuel theft if any.
Asset mapping for digital twins will be able to track and predict energy consumption at the site against their asset register, helping towercos avoid under-billing and to detect any additional equipment that was installed at the site by the MNO without their permission.
Effective data capture solution for digital twin modeling
Now that it’s clear which data can be prioritized for modeling digital twins, capturing this data becomes the next step in this process. Authenticity and accuracy of the data-being-captured are imperative to generate reliable digital twins, and data-capture-solutions aim to do just that to provide a comprehensive model of the external and internal condition of the towercos’ infrastructure and assets.
The use of cranes, which have traditionally been employed for data collection and tower inspection procedures by MNOs and towercos, was not only an expensive exercise but also came with a host of security and regulatory compliances that needed to be adhered to. This time-consuming process became even more challenging in high-density urban areas where towercos resorted to field-staff climbing up the towers to carry out site surveys and to collate data for preventive maintenance.
Drones are a welcome change that brought with them a host of benefits such as reduction in climbing and insurance costs while providing the potential for scalability and standardization in data collection.
Mat Jones, COO of Amplitel, in a recent panel discussion on digitalization stated that they were able to cut climbing costs by 30% a year by using drones intermittently.
Drones with their smaller sizes and capability to navigate hard-to-reach areas can provide multiple datasets from multiple visual angles to model a detailed digital twin. When equipped with high-precision cameras and sensors, they can identify and locate specific assets and even recognize problems such as rust or breakage.
Once towercos reach economies of scale, they will also be able to increase the frequency at which the infrastructure data is captured, against a current frequency of once every six months, which will pave the way to sustain the data needs of a reliable digital twin model without sacrificing the attention to detail.
Software specifically designed for site audits and tower inspections, helps the drones perform better across a wide variety of verticals. This software with its understanding of the tower industry and infrastructure can help program the optimal 3D flight plan to navigate complex field assets for quick and accurate data capture. Where upgrades of hardware can be expensive and time-consuming, the software provides flexibility in terms of scalability and modification based on the use case thereby driving better customer experience.
Some software vendors also offer a complete solution for MNOs and towercos that also include the hardware and the manpower required to inspect multiple assets across multiple sites.
This data captured holds immense potential to fuel the data transformation strategy that towercos need to create functional digital twins.
Process improvements with digital twins
Once these digital twin models have been generated, they need to be available in real-time to decision-makers who can, not only, view and analyze site parameters as a whole but also inspect individual towers and assets based on their location.
With the digital twin model in place, it becomes easier to make reasonable predictions by analyzing historical asset-related data to detect patterns in asset failures and maintenance timelines and to look for improvement opportunities in terms of reduction in downtime and capital expenditure.
For efficient process support, multiple digital twin models of an asset can be put in place to check for different variables and how the change in one can positively or negatively affect the other. The relationship between different processes, for example, how the records of site power outage not only have a direct bearing on the energy bills but also the fuel consumption and asset lifecycle of a diesel genset that goes into action during an outage.
Digitalization of infrastructure and asset records makes information more accessible at all levels of management across geographies, wherein data and decisions can be easily analyzed and explained. This helps in streamlining the management and maintenance workflows thereby making various operations and processes effortless. However, as we mentioned in our previous blog, the benefits of a digital twin system are incomplete without the backbone of a comprehensive site management platform that holds site and asset data.
A telecom site management platform to host your digital twin
The role of a telecom site management software becomes indispensable in hosting and benefiting from all that a digital twin has to offer. Visualization of data, as important as it is and as easy as it makes the data interpretation seem, is of little worth if present in silos. The ability to centrally index, search, analyze, and use the tower data in real-time from anywhere around the world within the IoT architecture is the essence of a digital twin which is only possible with a site management platform.
Software vendors that have been in the telecom space long enough to understand well the tower industry and the various players in it, such as the towercos, MNOs, and third-party service providers, and have partnered with other companies in the space based on their strengths to deliver data-capture-solutions, can bring in their expertise to stitch the data captured with the processes that are followed within these organizations to offer a solution with an added layer of artificial intelligence that can streamline the inevitable digital transformation and subsequent adoption of digital twins.
Within a telecom site management software, the digital twins sit at four different levels of advancement:
First Level - Repository of site data
The first level is basically a substantial repository of site-related information that lives within these digital twins, also known as the informative twin, which is uploaded onto the platform from thousands of tower sites across geographies in real-time for analysis and decision-making.
Second Level - 2D representations
The second level, or the descriptive twin, holds the two-dimensional representative engineering drawings that tools like AutoCAD, etc. make feasible, of the towers and the antenna load that they hold. This data is stored within the platform not only for the existing towers but for the towers to be.
Third Level - 3D representations
The third level is where towercos leverage drone software applications to have a more realistic and accurate three-dimensional view of the telecom infrastructure and assets at the site. This reality model view is where the telecom site management software providers are transitioning towards and are hoping to achieve in the very near future.
Fourth Level - Intelligent twins
Once your digital twin is in place, data from the model can be used to make critical decisions when a certain data point is not available in real-time. An intelligent data twin, or an autonomous twin, is a data-driven learning system that uses tower data from multiple data points to bridge information gaps wherever possible through the use of artificial intelligence, machine learning, and deep learning techniques.
Multiple twins of the same asset can also be created with process and time variables, which would allow for comparative analysis of these models in real-time to either pick more efficient processes or to detect, log, and predict changes in infrastructure and assets over time.
A comprehensive platform that can also be used to trigger inspections and audits makes it easier to collect and analyze the infrastructure and process-related data to take quick and effective decisions by providing a single source of authentic data to all stakeholders.
Providing standardization and open formats to both towercos and MNOs for data-sharing and integration can provide enhanced value-addition and flexibility in the way these digital twins are utilized while increasing the customer experience.
The future of data management strategy for tower owners
AI and ML-backed digital twin models are only possible in the abundance of reliable historical data that these models can train on which is currently lacking in the case of the tower industry. Most of the tower-related data reside in offline sources where it is either centrally inaccessible or corrupted making it more imperative than ever to focus solely on digitalization at scale.
The towercos that haven’t gotten on the digitalization bandwagon can start with assessing what site-related data is available to them and where it is stored. There should be a system in place to check the veracity of data and to look for data corruption if any that could have taken place when it changed hands during the acquisition, sale, or leaseback of towers. Only authentic, accurate data has to be picked to be fed to a digital twin model and data-capture solutions to be engaged to fill data gaps.
Even though the towerco data lake is still not ready to make the most of all that digital twin technology has to offer, this capability will grow in the years to come. Digital twins will stay latent until digitalization is not adopted as a whole by the tower industry.
However, the towercos and MNOs that have started digitalizing their structural and equipment data even for short-term reporting and decision-making will find it easier to achieve the potential that the digital twin technology has to offer. They will have the upper hand in the coming 5-10 years when the digital twin tech will be in full bloom because they will have reliable baseline data to model their twins on.