By Clément Grégoire, (XXX), Sofrecom
The digital twin really catches the eye! The idea of the digital twin concept is to have a digital model reflecting a real system, in real-time, and to provide the model with the automated functionalities of calculations, predictions, management for marketing, holdings, maintenance, etc. This concept is often associated with a 3D visual representation – as referenced in Building Information Modeling (BIM) – and its direct application to the Telecom industry can raise certain functional issues as well as business concerns.
Originally, a 3D Visualization Model
The 3D BIM model, designed for construction needs, facilitates the operation of complex systems where a lot of maintenance is to be considered. It was developed to address problems of planning between a multitude of actors and to harmonize the shared information on a single set of plans, thus improving the coordination of different vendors and departments during construction. In all phases of design, construction and operations, it centralizes the collection of essential information for business functions such as energy, structural calculation, conflict detection, compliance with standards, budget, deadlines, planning of intervention and more.
Is This Model Adaptable to a Network or to a Telecom Infrastructure?
The control and optimization of the value chain can benefit from a digital representation allowing the automation of operable business functions. But which of these representations will be adapted to best optimize the management of infrastructures and the delivery of network services?
A digital twin is a virtual model designed to accurately mirror a physical object or system. The studied system – a wind turbine for example – is equipped with various sensors related to vital functional areas. Beyond the visual aspect alone, the precision here applies to the vital parameters on which the representation must be based and through which all service functions must be planned and executed.
In the telecom sector, these parameters are the database of network or infrastructure elements with their interconnections and dependencies as well as the logistics necessary to manage services and their evolution (construction, operation, maintenance, marketing, QoS, etc.)
The nature of the form of representation (3D visualization or other) must be guided not by the problems of the BTP whose 3D representation is necessary, but by the identification of the vital functionalities to be collected: which parameters of the active and passive infrastructures (resistance, temperature, capacity, consumption, etc.) will we have to monitor? How often should it be done to help improve the construction, operation, maintenance and marketing of the network services and infrastructure?
Example of a Telecom Tower
If we take the case of a telecom tower, the vital information can be:
- Equipment inventory
- Available locations
- Energy consumption
As soon as the sites are designed, we can anticipate the methods that will allow us to identify these elements in an optimal way. The digital twin can be configured from the design plans of the site to be added, and connected to the digital model of the system. It is not necessary that an optical capture is made; what is important is that the model reflects the reality on the ground in terms of the information considered vital to the optimized management of the network and infrastructure.
Even if the site is already built and we do not have access to reliable plans, we find that 3D image capture and post-processing only comprise a small part of these "vital" needs, due to the nature of the information to be collected and the ability to measure it by image processing.
3D Image Capture Does Not Yet Fully Capture Vital Information
Today, the trend for the management of networks and infrastructure by digital twins is to automate the recording of real elements and their updating by optical captures. This offers virtual 3D visualization but poorly feeds the information vital for business functions. The question is, which parameters can be taken from a photo to be used in a calculation and which cannot?
Current Image Processing Features Are Not Mature
Image processing features allow the automatic detection of some equipment (e.g., antennas) or corrosion detection. However, reviewing these few elements by drone and post-processing is considered incomplete today and less advantageous than an audit by two technicians equipped to collect "vital" information onsite.
Generally speaking, capture and post-processing companies offer the service of drone or 360° capture, post-processing to build a 3D visualization and access to a library of features. This service is limited, however, due to the sheer number of elements to be recorded and deduced through image processing alone.
A drone service capture on a tower site and its post-processing today costs, on average, more than €1,000 with an average turnaround of about one week. This combination of capture and post-processing unfortunately does not allow for real-time execution/information.
If the digital twin depends on the drone capture, this means that it can only be updated between the period of capture and the next modification of the site (automated or not).
The Issue of "Real-Time" and the Automation of the IS Update: Necessary Inputs for the Vital Functionalities
Presently, the evaluation of the status of the site(s)/network(s) in real-time is still a dream.
It is easy to take readings on active equipment (e.g., readings, feedback), but how can this be done for passive equipment?
On certain functionalities, AI makes it possible to predict the evolution of the state of passive infrastructures by recording and assessing a certain number of past states and thus feeding predictive scenarios. However, the cost of such data acquisition can be very high (e.g., repeated drone applications) to regularly capture the whole of the infrastructure.
How to Consider Passive Infrastructure in Digital Twins
Are such passive infrastructures by nature built-to-last and therefore not requiring such real-time IS updates? Or can we apply these captures and feeds to a digital twin? ? If yes, how do we "activate" the process and how frequently do we?
In this framework, IoT is the exact tool to select the right sensor(s) to measure the vital information and feed it back to the digital twin.
There are, however, still cases that can be examined for optical capture and 3D reconstruction.
The Case of Flat Roofs
Without calling into question the nature of the vital information to be collected, a 360° capture of the roof during a technical visit in the design phase has a lower cost (~500€). This capture provides a visual model of the site that can be viewed at any time, to make distance measurements, automatic extractions of terrace plans, etc.
In those cases where we do not have the plans of the existing structure, such optical capture can be informative. Thankfully, the functionalities of distance measurements and production of automatic plans .dwg and .pdf are mature.
Reducing the Number of Site Visits to Obtain Missing or Inconsistent Information in the IS
Whether, checking a serial number, measuring the dimensions of free space for new equipment or reviewing the logistics of a power supply, the 3D site visit makes it possible to avoid certain superfluous trips. The key is knowing that these are indeed not marginal in the operation of the site or integrity of the infrastructure (optical connection node, mutualization point, towers, rooms, etc.). Further, is it necessary to redo all or part of the 3D visualization capture of the site in all instances that it is modified (frequency of modifications)?
The Case of Visual Representation in the Marketing Process: The Showcase Effect
The case can be made for the strong value of virtual visits to technical sites while in the process of renting the sites to operators and showcasing their value from afar.
For All These Reasons, It Is Difficult to Replicate BIM in the Telecom Sector
The fact that BIM has matured to today serve the needs of the construction industry, means that solutions already exist to meet the challenges of control and optimization valued by network operators. The introduction of new tools may prove too costly an addition given the budgetary constraints of the telecom network business functions.
However, it is relevant to continue to evaluate the progress of methods for capturing and recording "vital" parameters (IoT, sensors on drones, image post-processing, etc.) as well as their costs. Certain advances in sensors and post-processing could lead to the profitability of a capture service or an installation of an on-site system (IoT) if it responds to this automation of "vital" parameter readings and "automated" updates in the IS.