Introduction

At present, mining is immersed in the fourth industrial revolution, so there is talk of "Mining 4.0", in the framework of which technologies such as the Internet of Things (IoT1), Artificial Intelligence (AI2), Machine Learning (ML3) , Big Data Analytics4, Virtual Reality (VR5) and Augmented Reality (AR6), Digital Twins (DT7), among others, are applied, that is, it is the transformation of mining that we have been experiencing, through technological innovation in favor of converting it into an increasingly efficient mining in all areas (business, society, the environment, among others), in contrast to what Byrne & Engdahl (2021) have reflected, that there is currently 70% operating efficiency, between 30% and 50% of operating costs are spent on plant, fleet, and equipment maintenance.

If a mining that coexists in harmony with the communities that could be in the area of influence of the operations, that reduces emissions, that reduces or eliminates the negative impact on the waters, that really protects the safety and health of its workers, that facilitates the prevention, monitoring, and control efforts of government agencies, and that increases the benefits of shareholders, is desired, then it is necessary to implement Mining 4.0.

Mining 4.0 is possible, to the extent that more and better data is constantly being collected, here sensors play a key role.

Sensors

Kiziroglou et al. (2017) stated that the use of sensors is already widespread in mining, including in applications such as automation and remote operation, and data analysis for control and optimization, they have also stated that in mining, sensors are of great relevance in the control of driverless vehicles, the remote operation of equipment and also the management and monitoring of assets, site security and the location of personnel, for this, the main sensors are satellite positioning systems (GPS) and positioning systems (based on radio frequency), dead reckoning systems (position tracking based on inertial sensors), beacon sensors and tagging technology.

Kiziroglou et al. (2017), complemented that imaging, 3D ranging and mapping sensors are widely used in local and remote surveillance of operations, monitoring, and control of equipment and vehicle automation, in large-scale surveying and mapping, battery monitoring of storage, monitoring of the evolution of excavations, tonnage, rock monitoring and particle measurement. These sensors can be: CCD8, CMOS9, infrared, depth; also, three-dimensional mapping, which is based on a combination of optical and position/orientation sensors, sometimes complemented with other hardware such as depth and proximity detection, and spatial reconstruction software.

In relation to proximity sensors, Critchley (2019) has pointed out that they are installed in machinery, vehicles, and equipment to avoid collisions of any kind and have devices that emit visual and sound alerts.

In the activities of quality control from operations, Kiziroglou et al. (2017) revealed that MineSense has brought to market systems capable of detecting payload, including ore grade assessment, using high-speed XRF10 and high-frequency electromagnetic spectroscopy. Its ShovelSense product, which can be retrofitted to existing shovels, is capable of deciphering ore tailings in real time, in addition to quality analysis. This information, provided to the machine operator, allows the sorted ore to be optimally loaded onto waiting trucks.

In exploration, sensors are used in remote sensing methods in order to cover large areas of land and discard areas, in such a way, to concentrate on those that indicate anomalies, in this regard, Kiziroglou et al. (2017), mentioned Gravity Gradiometry and Time-lapse Seismic monitoring.

One of the uses of sensors in underground mining is when Automated Temporary Roof Supports, Mobile Roof Supports and Automated Roof Bolting Systems are implemented, and then sensors can be used to measure if there is any movement within the rock face over time, and to inform operators if further measures need to be implemented. One of the most innovative uses of sensors in the mining industry in recent years is the shift towards automated mining operations. While this has been aided in conjunction with the implementation of IoT, Industrial Internet of Things (IIoT), and advanced data analytics methodologies, sensors are still at the heart of data collection processes. (Critchley, 2019).

“The use of smart sensors tends to generate 34 billion dollars in value for the sector, because it facilitates the maintenance and use of equipment, reducing failures and reducing downtime.” (Carvalho, 2021).

The technology of sensors placed in the drills, both for self-diagnosis and for gathering information on drilling variables, geology, and the quality of the rocks being drilled, contributes significantly to decision-making in real time.

Sensors are very useful for measuring the composition of materials being processed at different processing stages, according to Kiziroglou et al. (2017), they serve to control the benefit equipment and the classification by grade of the ore during processing, also in the excavation and hauling stages.

Kiziroglou et al. (2017) have indicated that XRF and X-ray diffraction (XRD) are used for ore grade analysis in the laboratory setting, but on-site portable instruments for some analytical techniques are becoming increasingly available. They have also argued that advances in micro-engineering have enabled the development of portable devices such as the miniature Mossbauer spectrometer, originally developed for NASA and now proposed for mining analysis.

Complementing what was stated by Kiziroglou et al. (2017) in terms of metal detection sensors, there is the Magnetic Resonance sensor technology presented by Next Ore (2020) with which it is possible to detect metals that are being searched for.

They explain Next Ore (2020), that Magnetic Resonance (MR) is a form of Radio Frequency (RF) spectroscopy that can be used for the quantitative measurement of minerals in ores. In the MR technique, each mineral has a specific “resonance” or response at a particular radio frequency. Resonances are highly discriminating, as it is extremely rare for a resonance to overlap another mineral's resonance. Therefore, the method involves tuning one or more specific minerals of value to the target operation, all of which enables high-speed sorting of bulk minerals; has been proven in various operations around the globe with varying throughput rates in both passive detection and active bulk ore sorting systems.

Kiziroglou et al. (2017), have exposed that the integrated analysis of data from heterogeneous extraction and transport sensors, with data from sensors generated in the processing (such as from air/gas mass flow meters that can be beneficial to improve flotation efficiency during the foam phase and the hydrodynamic characteristics of the pulp phase), could lead to more accurate and efficient end-to-end management of products and systems.

Hansford Sensors (2021) have argued that low profile, side entry accelerometers can be used for crushers, screens, conveyor pulleys and drives, and where access is restricted, and that these sensors are suitable where debris or dust can affect the performance of a device.

Robben & Wotruba (2019) cited by Nwaila et al. (2022), described the development in sensor-based dry mineral classification and concluded that this type of classification is receiving more consideration within the minerals industry. (See table 1). 

The use of sensors for health and safety, according to Kiziroglou et al. (2017), can allow continuous monitoring of the location and health status of mine workers, which is possible thanks to portable sensors that measure parameters such as the level of cumulative exposure to hazards (for example, radiation or powder), heart rate and blood oxygenation, among others. The collected data is transmitted in real time to an analysis server.

Complementing what was stated by Kiziroglou et al. (2017), Critchley (2019), in terms of safety, has mentioned gas, dust, tiredness, and fatigue sensors.

Regarding dangerous areas, Kiziroglou et al. (2017), have reported that in hazardous area fencing systems, portable detectors can provide a warning and trigger a preventive shutdown when approaching or entering an unsafe area.

In terms of sensors, the safety helmet presented by Pradeepkumar et al. (2021), which consists of a “smart” helmet based on LoRaWAN11 technology and that contains a device that keeps the user alert about the quality of the air in their environment, so that user can take measures to avoid damage to their health.

Internet of Things

The IoT, given its great capacity to collect information in real time, as well as its ability to process, correlate, and analyze it, represents a great resource in operations, as well as in the monitoring and control of plans; data related to personnel, machinery, equipment, tools, third-party services, among others, can be managed in favor of productivity, all this, within the mine facilities or anywhere on the planet with adequate internet service.

It is a platform for the integration of internal and external activities at the mines, which are part of the operation, as well as clients and suppliers, as long as there is integration of information. Going further, it can also be the appropriate means for the integration of the different mineral resource production operations that the same firm or corporation may have in different parts of the planet, being able to achieve economies of scale and synergies that are more difficult to achieve separately or cannot be achieved with maximum efficiency in the interest of maximizing profitability.

The deployment of 5G cellular network technology contributes significantly to IoT processes, AI, as it is capable of greater bandwidths, higher speeds with lower latency (around 10% less than that obtained with 4G networks) all this will result in greater information capture (more devices connected and exchanging data) and its processing, as well as the use of the “cloud” to store, will also favor the implementation of DT, with which simulations and evaluations will be possible, in constant evolution, thanks to the possibility that the mathematical models involved are constantly improving from the data that is obtained in real time.

Molaei et al. (2020) have stated that mines are using IoT for gas detection, machinery positioning, personnel location and tailings dam monitoring, and have indicated that it is a technology that looks promising to achieve in mines, that the consumption of energy to be more efficient, to create intelligent environments, to optimize risk management and mitigation plans, to use the “cloud” as an information store, and for “smart” production, among others.

Carvalho (2021) stated that the IoT can help reduce risks during the processing of coal mixtures, control the unloading zone of the mines and eliminate possible errors caused by human error, which reduces the costs of the process and the service, and increases the quality of products.

Based on IoT, preferably with 5G networks or failing that, any other low latency option, and with the contribution of VR and AR technologies, mining is tending to establish Remote Operations Centers (ROC), which have a very important value for the safety and environment of work for employees, since it makes it possible to extract them from places as risky as mines, regarding these ROCs; Cornejo (2022), has expressed that they help promote the transfer of knowledge among the miners, since they are all in the same place at the same time. Other benefits include a reduction in the costs associated with moving both qualified specialists, and mine workers to the site.

In favor of occupational health and safety, diverse technological resources can carry out dirty, dangerous and degrading jobs that were previously carried out by workers, in addition, AI is used to carry out medical diagnoses of workers. (Cornejo, 2022).

Carvalho (2021) has exposed that through technological innovations, it is estimated that by 2025, worldwide, a thousand lives can be saved, and 44 thousand people will not be injured during mining operations.

Vella (2017) cited by Hyder, Siau & Nah (2019) have indicated that with the help of AI, ML and autonomous technologies, the exposure of workers to hazardous mining operations can be minimized. The machines can autonomously monitor the atmosphere, send signals and warnings, locate problem areas, and work continuously even in dangerous situations.

Cornejo (2022) argues that digitalization and the implementation of autonomous technology have had a great positive impact on the health and safety of miners, in many ways. The most important factor is, of course, removing the operator from the cab of the machine and allowing them to operate from the safety and comfort of an ergonomically designed chair. This also eliminates exposure to noise, fuel particles, and extreme weather conditions. In addition to this, operators are protected from possible dangers such as machine collisions or falls while carrying out tasks on site.

Hyder, Siau & Nah (2019) have expressed that data visualization and analysis techniques can be used to analyze the causes and factors that lead to accidents and preventive measures can be designed with a greater focus on eliminating the causes of accidents. Intelligent systems designed with a focus on eliminating potentially hazardous situations, reducing or completely eliminating human presence in risky and hazardous work, installing roof supports, and removing hazardous gases and dust, can help reduce accidents and deaths.

Artificial Intelligence and Machine Learning

Marr (2016), and Wang & Siau (2019) cited by Hyder, Siau & Nah (2019) have agreed that AI and ML are two technologies that have the potential to change the technological framework of the future and both depend heavily on Big Data Analytics. All this has been complemented by Ali & Frimpong (2020) who have stated that AI and ML are currently the two main pillars of modern automation technology, and that these two fields have influenced almost all industries with thousands of millions of dollars invested to advance beyond the limits; they conclude, stating that in mining, they can be applied from the beginning of the mining project to the end of the mine life cycle, from prospecting to production, closure, and mine reclamation.

AI systems and data analysis software can be fed geological, topographical, mineralogical and mapping data and used to identify anomalies and variations in the data and to locate areas of potential interest. Some research in this field is already underway, and Goldspot Discoveries Incorporated is using such a system as an experimental basis for gold discovery. Similarly, Goldcorp, and IBM Watson are working together to filter big geological data to improve the accuracy of mineable prospect selection. (Walker (2017) cited by Hyder, Siau & Nah (2019)).

AI and ML can be applied to develop stand-alone drills that can locate potential sites identified in the prospecting stage and conduct drilling activities, and can feed drill log data into the system. This technology can also be implemented during production drilling. (Hyder, Siau, & Nah, 2019).

Di et al. (2019), and Ghorbani et al. (2020) cited by Nwaila (2022), have shown that there are many good reasons to apply Data Analytics and ML in the mining and metals industry, such as: (a) the availability of vast data sets that they often cannot be visualized and interrogated practically; (b) complex systems that may exclude fully deterministic and reductionist solutions; (c) prediction of geological structures/features and performance of metallurgical plants that cannot be reconciled using conventional methods; and (d) a need for business improvement.

Kobold Metals has been employing AI and ML to find mineral deposits, while arguing that the mining industry is poised for a major digital transformation; among other activities, they have been compiling scientific and historical data, among the information that is in the public domain, of which 20 million pages are estimated, among which there is what they have called “dark data” (that has been forgotten, or not have been used), and that by being digitized and processed with the technology available to them, among which spectral satellites stand out, they can help build a kind of Google® Map of the earth's crust; with all this, Kobold Metals aims to increase the success rate to 20% in contrast to the scenario observed in the last 30 years in the world in which “the number of discoveries per dollar of exploration capital has been reduced six times”. (McGee, 2022)12. 

According to Okada (2022), Convolutional Neural Networks13 (CNN, the acronym ConvNet is also used) and Deep Learning (DL14) are expected to effectively contribute to the achievement of geophysical data inversion.

Ali & Frimpong (2020) have shown that the use of surface and subsurface images can be implemented for mapping and geological exploration, relying on state-of-the-art DL algorithms together with CNN, with which high precision can be obtained.

According to Ali & Frimpong (2020), AI can be used to achieve smart planning and operations in mines by making use of “advanced master data management, workflow management and predictive analytics capabilities to analyze different types of data in real time, and optimize short-term planning, scheduling and transportation”; they have also argued that intelligent algorithms can be used to support the decision-making process aimed at selecting the production assets required in a mining project.

Hyder, Siau & Nah (2019) have explained that an important application of AI in the mining industry can be the detection of hazards, especially dangerous gases, toxic dust and radiation in the mine. AI systems can be developed to inspect the workplace before the workers through the use of robots, sensors, and data collection from pre-installed monitoring stations. These stations can activate alarms, give warning signals and block off the affected area to reduce the further spread of the hazard. When connected to mine fans and ventilation networks through intelligent systems, AI systems can direct airflow, increase or decrease the amount of air and pressure of mine fans, and turn on and off stop certain fans to automatically direct hazards out of the mine. This can improve the safety of mining operations, reduce downtime, increase productivity, and lower accidents, and related costs.

Hyder, Siau & Nah (2019) have predicted that AI and ML can be implemented in machinery such as shearers, coal cutters, jumbos, conveyors, cutting heads, and road headers to direct their operations, automate the application of energy in the cutter heads to match rock strength, and hardness, monitor gas and methane inrush during operation, continuously monitor the condition of the roof while it is in operation, and disseminate data on working conditions to make informed decisions, and take corrective action far before the problems escalate.

Regarding sampling, Hyder, Siau, & Nah (2019), have predicted that autonomous samplers can sample minerals, atmosphere, gases, dust, and toxic materials, even in areas of high concentration. Intelligent continuous monitoring systems can provide early warnings, suggest preventive measures, and reduce the need for workers to access the hazardous area to take samples, and these AI systems can reduce the need to bring in samples for laboratory testing by providing images, perform on-site tests and communicate the results when necessary.

Hyder, Siau & Nah (2019) argue that the dangerous aspect of roof support in underground mines is automatable as an integral part of continuous mining machinery.

Regarding the transformation of minerals, Martens et al. (2021), based on what was stated by different authors, between 2005 and 2021, has stated that Computer Vision15 is used for the analysis of size distribution, mineral classification, mineral material composition analysis, digital image processing in metallurgical plants and froth flotation analysis, have also stated that advances in technology, such as in-line soft sensors for mineral processing, and ML algorithms, empower new ways to optimize grinding and flotation processes.

For both crushing and classification, both AI and ML are useful for intelligent and automated control of particle size monitoring and mineral content. Chauhan et al. (2016), Okada et al. (2020), and Deo et al. (2021) cited by Mishra (2021).

In terms of crushing, Tessier, Duchesne & Bartolacci (2007) cited by Mishra (2021), have argued that sensors and AI (SensAI) go hand in hand. The goal is to optimize the performance of the crushing stage. For this, the sensor typically needs to detect the particle size or mineral signature of the particles at both the inlet and outlet of this step. Sensors with ML algorithms can be used to ensure that the size of rocks entering a mill, and onto the conveyor belt is not too large to cause failure; additionally, Zhou & Sun (2020), and Ostasevicius et al. (2021) cited by Mishra (2021), have stated that SensAI can also be used to monitor the wear of crushing and sizing equipment.

The speed of execution or computational complexity of ML algorithms is of critical importance, as most decisions need to be made in real time as minerals move on a conveyor belt. If the features are carefully selected, simple ML algorithms can also give very good results. (Mishra, 2021).

In the case of the concentration phase, Mishra (2021) has stated that the methodology involves the use of sensors (sometimes existing and sometimes new) to generate data about the process, which are then used by ML algorithms to optimize and diagnose the process (in search of possible failures), on the other hand, Horn et al. (2017) cited by Mishra (2021), have recommended that for float cases the use of CNN is preferable given the low performance obtained with ML.

In dewatering, ML-based algorithms have shown potential to improve efficiency. (Raman & Klima (2019), and Tripathy et al. (2021) cited by Mishra (2021)).

According to Hosseini & Samanipour (2015), and Feng et al. (2015) cited by Mishra (2021), sensors and AI are increasingly used to obtain and process information about the characteristics of materials, at the beginning, and at the end of each stage in a processing plant.

According to Jeswiet & Szekeres (2016) cited by Hyder, Siau & Nah (2019), AI-based systems can be designed to classify minerals. These systems may use color sorting, X-ray transmission, or near-infrared sensors to remove debris from the ore. They can be designed to take advantage of differences in physical properties such as specific gravity, density, gloss and weight, mineralogical compositions and chemical properties. Applying these systems before grinding and crushing equipment, can greatly increase the efficiency of the crushing process and reduce energy cost, as crushing and grinding are the most critical parts of the mineral processing cycle, more energy consume, and are less efficient.

Ali & Frimpong (2020) have argued that being able to predict the concentrate grade and recovery for a given configuration, and obtain optimal values for all important variables involved, is key to designing an efficient flotation configuration for coal and complex metallic minerals. ML and AI can help design an algorithm/model that, based on the plant goal, includes optimal flotation conditions involving all significant variables, and can maximize the plant goal.

Mishra (2021) has argued that as ore processing increasingly requires more chemicals that are more tailored to the material at each stage of the process, this is a rich field for AI innovation.

ML and AI models can potentially complement existing process models in mineral processing. Because these models are faster to develop and easier to refine, they can prove valuable in mineral processing. (Eirinakis et al. (2020), and Zobel-Roos et al. (2020) cited by Mishra (2021)).

Nwaila et al. (2022) argue that sensor technology, data processing, analysis and image recognition based on AI are fundamental requirements in the short and medium term to create autonomous discrimination and classification systems that could include functionalities such as sensor-controlled classification using mechanical separation.

The use of sensors and AI, according to Betrie et al. (2013), and Tousi et al. (2021) cited by Mishra (2021), may be used by government agencies to monitor the effects of tailings in mines, which may be possible thanks to remote sensing as stated by Hao, Zhang & Yang (2019), and Yan et al. (2021) cited by Mishra (2021).

Mishra (2021) has argued that as part of the evolution of mineral processing, the use of organic compounds will be necessary, and that AI has proven to be efficient in the design of the appropriate catalysts.

abcdust (2022) state that AI and IoT can help optimize dust control measures on haul roads, and crushers, ensuring the correct dose of additives and water is applied, at the right time and in the right way, for dust control. Not only will this improve dust control at the mine, it will reduce water, additives, energy, labor and associated CO₂ emissions by 50%. In the particular case of haul roads, it will also reduce the risk of trucks hydroplaning, and the frequency of road maintenance, avoiding accidents and loss of productivity.

Immersive technologies

The most used immersive technologies in mining are: VR, AR, and DT.

VR is generated by software and audiovisual transmission, it is about the recreation of aspects of real life; in these recreations, users feel inside them. The usefulness lies in the possibility of performing simulations, and for education.

It is possible with VR to take a virtual tour, even on a 1:1 scale, of a mine, take an “aerial” or “underground” or “surface” tour, you can see the geological structures, the drilling that has been carried out, the infrastructures that exist, and those that have been planned, as well as those that existed, the machinery, equipment, vehicles and any other that are in operation, the processing plants, and all representation of the mine that has been fed and is being updated in real time.

AR allows remote inspection and assistance by experts, who are remote, to machinery, equipment, processing plants, among others; it is also possible to carry out maintenance and repairs assisted by experts, who, being remote, can guide less experienced personnel in carrying out what is necessary, can send people in the field, manuals, images and comments, can guide them step by step.

In terms of operations, AR is very useful for supervisors and experts, without having to be in the mines, to guide operations, monitor and control, and prevent accidents. This technology has potential in this area to increase productivity and could encourage the commitment of workers with their obligations and organization.

Through DT technology it is possible to have a virtual copy of the different mining processes, in this copy it would be possible to test changes and evaluate results before implementing them or deciding not to implement them, of course, precise digitization of absolutely everything that is you want to replicate, it is required to apply the aforementioned technologies and others that will be mentioned later, if required; the effectiveness of this technology depends to a large extent on the digital base information of land, machinery, equipment, vehicles, tools, various production infrastructures, and in short, everything related to what is wanted to be represented, on the other hand, is highly important to have the appropriate quantity and quality of sensors that are constantly collecting information about everything that happens, undoubtedly, all this without adequate mathematical modeling and its versatility to undergo modifications based on the processed data and the results generated, it would not be really helpful.

Creating DT and virtual models and roadmaps requires new and unique skill sets that include critical thinking, data visualization, and proactive decision-making based on predictive analytics. (Deloitte (2021) cited by DCC (2022)).

There are numerous opportunities to optimize operations using DT technology, from pit to port, explained Martin Provencher, Industry Director, Mining and Metals, AVEVA, as reported by DCC (2022).

Singh et al. (2021) have expressed that there is a high cost associated with the implementation of DTs, which undermines the aspiration of reaching the maximum potential of this technology; have exposed that the whole process of developing ultra-high-fidelity computer models and their process simulation to create a DT is a time- and labor-intensive exercise that also requires a great deal of computing power to run, which makes that the DT is an expensive investment.

By means of AR, it is possible to contribute to the realization of better, more precise and effective perforations, with which it is possible to avoid the undesired results of blasting; it is possible with this technology for operators to interact with expert drilling and blasting engineers.

Leveraging DT technology for drill and blast designs and digital modeling of life of mine plans can help operators better forecast, allocate, track and manage resources. This can improve data management, accuracy, and efficiency. (Deloitte (2021) cited by DCC (2022)).

Regarding the DTs, Nad et al. (2022) have shown that metallurgical plant simulations allow DT applications to predict short- and long-term process operation scenarios with alternative operating actions, and varying mineral characteristics. It integrates mining data, feed ore characteristics, and plant online sensor data with a detailed dynamic processing model based on plant mineral particles, and includes powerful online prediction and adaptation algorithms for model parameters. The results of what-if predictions from a metallurgical DT are readily available to select the best targets and control strategies for advanced process control of different plant areas, find the best approaches for different production events and scenarios, and even estimate the missing information from the sensor.

Several commercial software tools are available to build models that describe the dynamic and steady-state behavior of mineral processing circuits. Some examples are USIMPAC, JKSIMFloat, SUPASIM, MODSIMTM, and HSC Chemistry Sim. (Nad, et al., 2022).

They affirm DCC (2022), that DT technology can be used to improve efficiency and reduce carbon emissions throughout the supply chain.

Another great contribution in the workplace is represented by VR, through which it is possible to train employees of mining companies, whatever their level in the organization, making training safer and more efficient, in such a way that those trained can fully practice what they have been taught, in a safe environment, before developing it in the real world, is to note that for this training it is not necessary that those who train are present in the same place, they can be anywhere in the world with adequate access to the Internet, and with the technology required for VR.

VR allows accidents to be reconstructed, safety training for workers and planning for possible emergencies that may arise. On the other hand, an expert in industrial safety could also see possible causes of accidents, long before they happen, if the representation of VR is faithful to that which exists in physical reality.

Immersive technologies in the mining industry have made it possible to create a mining metaverse which they have called Minverso (https://minverso.cl/), and which consists of thematic showrooms in which, using VR glasses, it will be possible to access from anywhere in the world, with adequate internet connection, in fact, an application is even being developed to access from the smartphone; these showrooms are designed to adapt to the needs of companies, they will be spaces for development and training.

It is noteworthy that there are risks for people in the use of immersive technologies such as, so far, they are: potential to create accidents due to distraction, also, by using VR nausea, eye problems, headaches can occur.

Remote sensing

Among the various remote sensing equipment we have the Intelligent Unmanned Systems (IUS), which are being very useful in different areas of the mining industry, they are very useful for topographical surveys, digitalization in underground and surface mines, measurements of progress of operations blasting, inspect operations, inspect access roads and haul ramps, slope monitoring, material pile inspection, environmental monitoring, inspection of dumps and deposits, transportation of tools and other artifacts, security inspection of mining properties, among others.

It should be noted that there are several terms related to these systems, namely, the term drone is used commercially, although this word is really referred to the flying equipment that in general terms can be an RPA (Remotely Piloted Aircraft) or a UAV (Unmanned Aerial Vehicle). Now, when it comes to the flying device, plus the link, plus the ground control, then the phrase “system” is used, thus existing the RPAS (Remotely Piloted Aircraft System) and the UAS (Unmanned Aerial System).

RPAs or UAVs can be equipped with RGB16, LiDAR17, multispectral18, thermal19 or thermographic sensors, or magnetometers20, among others, with which it is possible to carry out different information capture activities, both in surface and underground mines, to feed artificial intelligence systems.

Castro (2022) has indicated that in greenfield projects RGB images and photogrammetry may be sufficient, but in brownfield projects it is preferable to use LiDAR technology because greater precision is required in surveying existing structures, and he has stated that this technology is accurate even in adverse conditions such as underground mining where it is possible to capture the work in great detail despite conditions of little or no light, presence of dust and water. Conditions that would make photogrammetry impossible.

Currently, the most widely used aircraft of this type in mining are the Matrice 300 RTK by DJI, the Wingtra one GEN II by Wingtra AG or the Ebee GEO by Sensefly.

The information collected by UMS could be feeding the operations systems in real time, acting as sensors, and contributing to the improvement of the various operations, and contributing to decision-making.

Schodlok, Frei & Segl (2022) have exposed that hyperspectral remote sensing has been used in geology, soil, exploration, and mining. New hyperspectral satellite systems such as PRISMA (PRecursore IperSpettrale della Missione Applicativa) and DESIS (DLR21 Earth Sensing Imaging Spectrometer) are in operation, the European next generation hyperspectral satellite CHIME (Copernicus Hyperspectral Imaging Mission for the Environment) and the German system EnMAP (Environmental Mapping and Analysis Program) was successfully launched into space on April 1, 2022.

Hyperspectral satellite data is an important data source for exploration activities, despite moderate spatial resolution. However, for detailed mapping, especially within a zone of mineralization, high spatial resolution hyperspectral data from UAV or airborne systems are relevant to understanding and mapping the mineralogy of that zone throughout its entirety complexity. (Schodlok, Frei, & Segl, 2022).

Muñoz (2021) has pointed out that the exponential progress of digital technology, and the appearance of drones for photogrammetry, allow precise knowledge of the geometry of the fronts to be blown up, or of the sections of the galleries in the case of underground mining; he has also expressed that the use of drones is very significant in the digital design of blasting, and that the safety of personnel in the field has improved significantly, while a reduction in the time for obtaining data for blasting optimization has been achieved, and increased accuracy; on the other hand, he affirms that “drilling must also take advantage of the latest technologies to gain precision for the benefit of the client”.

Autonomous operations

Autonomous operations allow costs and accidents to be reduced, which is in line with the constant search for better productivity levels that help improve competitiveness and increase the value of shares.

Autonomous operations require digitization as precise as possible, which is very useful base information in the systems that have been described in the preceding paragraphs.

According to Adams (2022), to communicate, autonomous mining trucks rely on: wireless communications, object detection/avoidance systems, Global Positioning System (GPS); he has also stated that: “These trucks can transport up to four hundred tons of ore and transport it accurately without human interaction”. What Adams pointed out is part of what Kiziroglou et al. (2017) had pointed out about autonomous trucks and the sensors they use.

Additionally, Kiziroglou et al., had pointed out that the developments foreseen in this area derive mainly from the rapid development of Advanced Driver Assistance Systems (ADAS) for passenger cars, which include functions of navigation, collision avoidance, assistance of parking and automatic parking, lane change assistance and adaptive cruise control. Due to the variety of requirements, these systems also integrate ultrasound and LiDAR sensors.

Mine automation is expected to benefit from the improvement and cost reduction of these systems in the automotive market, including, not only sensors, but also related software for functional integration. The result will be a much richer combination of vehicle automation sensor data, and more precise and reliable control. The availability of data regarding the fleet will also allow the optimization of high-level processes, taking advantage of stochastic data analysis techniques for vehicle maintenance.

BHP and RioTinto use a fleet of autonomous trucks and have reported a 15% reduction in operating costs compared to manually operated trucks (Dyson (2017), and Simonite (2016) cited by Hyder, Siau & Nah (2019)).

Komat'su (2021) has reported that mining trucks under the control of its FrontRunner® technology have transported, worldwide, 4 billion tons of materials through 400 trucks; while Caterpillar (2022) has reported that by the end of 2021 the autonomous trucks manufactured by them with their Cat Command® technology had transported 4 billion tons and that this year more than 500 of these trucks circulate in the world. It should be noted that by May 2022 there were 1,068 autonomous mining trucks in operation worldwide, according to Global Data (2022), of which 706 are in Australia at 25 mines.

Crozier (2016) cited by Hyder, Siau & Nah (2019) has stated that at BHP Billiton, autonomous drills are equipped with sensors, inclinometers and other instruments to perform drilling tasks autonomously, and feed data to drilling analysis packages, data to further refine the drilling machine.

Ali & Frimpong (2020), regarding dragline automation, have stated that:

• a group of researchers worked on the development of a semi-autonomous dragline model to move 200,000 tons of material in more than 12,000 cycles (Corke et al. 1997, 2003; Roberts et al. 1999; Winstanley et al. 1999). However, each of those models failed to address a core common problem during the operation of any autonomous excavator, which is the automatic adjustment of a machine to ground obstructions.

• As research in the modern automation industry has shifted towards vision-based systems, some researchers used image recognition techniques to achieve better autonomy of operation. Chi & Caldas (2011) proposed the idea of employing neural networks and Bayesian classification models for object recognition on a construction site. Ji et al. (2016) demonstrated the application of a vision-based system for the detection of trucks and hydraulic excavators, with a success rate of 73 to 89%...

• … The most recent work on dragline automation was done by Somua-Gyimah et al. (2019) using the state-of-the-art deep learning approach with convolutional neural networks (CNN) for terrain reconnaissance and object detection tasks in mining excavation operations.

• Ali and & Frimpong (2020) about the advances in automation of machinery in underground mines have contributed the following:

• …the most advanced implementation of an autonomous system in the underground environment is LHD automation. The system navigates the mine by detecting tunnel walls (Roberts et al. 2000). The operation was demonstrated in 1999 and has been operated at various mines since its successful start and application.

Cat (2021) have presented their Cat ® Command for Excavating technology by means of which it is possible for excavators from 20 t up to 40 t to be operated remotely in two options, namely, with line of sight (up to 400 m) and without line of sight from the comfort of an office with access to a wireless network; this technology has been designed so that in cases where an operation could be risky for the operator (for example: after a blast), the operator is not present in the machine.

The technology also allows people with limited to get on the machine can operate it from outside. Similar to this technology, that same manufacturer has Cat® MineStar™ Command for Dozing through which the advantages described in its excavator technology are obtained, but in this case for chain tractors.

Autonomous operation of road trains is virtually a given by Mineral Resources Limited in Pilbara Australia in partnership with specialist automation company Hexagon for Mineral Resources, this is the solution to be able to extract the existing iron ore in the Ashburton Hub project, which to date cannot generate profitability by any other means other than this autonomous operation. Until now, tests have been carried out with trucks of three trailers to transport up to 385 tons with a goal of 425 tons; up to 5 trucks with three trailers each are expected to operate; all of which will be possible thanks to drive-by-wire technology and an autonomous management system.

Regarding the drive-by-wire technology, Parsania & Saradava (2012) have stated that it is a technology of electronic systems that replace the old mechanical controls. Instead of using cables, hydraulic pressure, and other things that give the driver direct physical control over the speed or direction of the vehicle, drive-by-wire technology uses electronic controls to apply the brakes, control the steering, and operate other systems.

These trains will be operated from a remote control center, from where a driver will control the first truck in the fleet, the others will have no drivers, the unloading of the trailers will be fully automated.

In terms of rail transport, in 2019 the first automated railway network, called AutoHaul™, was put into operation, which can transport one million tons per day and which contributes to the reduction of accidents and costs, and “improves the times of cycle by using information on the topography of the network to calculate and offer a safe and consistent driving strategy”. (Rio Tinto, 2021).

Autonomous drilling technology allows a single operator to manage multiple autonomous drilling units for greater precision and productivity.

Cat (2018) stated that through its Cat® MineStar™ Command for drilling technology, it is possible to remotely ensure that all holes are drilled precisely in the right place, to the right depth and at the right angle.

Rio Tinto's Gudai-Darri mine in the Pilbara-Australia has the following advanced technologies (Mining.com, 2022): robotics for the mineral sampling laboratory, as well as for the distribution of parts in the new workshop; autonomous trucks, trains, and drills; a complete digital replica of the processing plant that allows teams to monitor and respond to data collected from the plant, the same digital asset data is used to provide an interactive 3D environment for virtual reality training. These autonomous assets are monitored remotely from Rio Tinto's operations center 1,500 km away in Perth.

The Vale company has 72 autonomous vehicles, in mid-June 2022, and expects to end the year with 86; the use of autonomous trucks has brought the company a 25% increase in the useful life of tires and engines, they have also been able to increase the speed of the trucks from 40 km/h to 60 km/h, which has produced an increase in productivity of 10%; additionally, in the Carajás yards, where it has 12 autonomous machines, it has managed to reduce operating deviations in the formation of ore piles by 90%, which generated agility in train loading. (Diniz, 2022a).

The European Union implemented the X-Mine project which was based on XRF technologies, X-ray transmission (XRT), 3D vision and its integration with mining and mineral classification equipment. The project was implemented in 4 mining operations from small to large scale, and different metals, including some of those considered critical metals. The detection technologies developed in the project improve the efficiency of exploration and extraction, resulting in fewer blasts required, it also allows “more efficient and automated mineral selectivity in the extraction stage, improving the options of previous concentration of the mineral, and resulting in less use of energy, water, chemicals, and… worker exposure during downstream processing.” (CORDIS, 2022).

Conclusion

Although this is a brief review of Mining 4.0, it is possible to conclude that this evolutionary phase of mining is here to stay and continue the necessary process of improving the industry. The benefits that can be obtained are very evident, for entrepreneurs, society, environment, and governments.

There are those who have decided to speak of “intelligent mining”, but perhaps it is more appropriate to refer to the mining that is desired, such as Holistically Harmonic Mining, as already stated in Paulo & Herrero (2021).

In any case, Mining 4.0 is a logical step for current times, however, its implementation by any company of any scale is not easy, this step is expensive in terms of investment, and represents a true transformation of the industry, which aims to be a positive transformation that will lead the sector towards greater efficiency, less negative environmental and social impacts, increase in the value of the companies' shares, however, the implementation process will provide information about the negative aspects that they could be generated, and will surely be duly attended to for their solution and improvement.

Everything indicates that companies with less economic possibilities should begin the transformation towards 4.0, little by little, but to start, it would be necessary to detect the areas in which it is less expensive to start the process, and it could generate the greatest amount of benefits at the same time. It is evident that the 3D vectorization of mines and the integration of such vectors is a good start, since this information is the basis for everything else that has been described here.

It should be noted that many mines, perhaps the majority, already have 3D vectors of their deposits, their mine designs, their various infrastructures, in short, what we have to do is start to integrate, and build the rest of the vector information that is not yet available, probably with any of the sensor technologies that have been mentioned in this paper; then assuming that there is adequate internet service, it would be necessary to hire a reliable “cloud” service, since that is where the large volumes of information generated in a mine will be handled, and all this data is confidential to the company, in such a way that it is necessary to “guarantee” that the information will not be hacked; apparently, everything else consists of hiring suitable professionals for these levels of technology that have been described, training existing personnel and hiring service providers for the different technologies that are required in the framework of Mining 4.0.

Not having modern and reliable internet services in a country, or in some mining region, will make it “uphill climb” to have better mines.

How to efficiently carry out the implementation of Mining 4.0 in different types of mines, of different scales, located in different countries with different realities, represents a topic for applied research.

References

abcdust. (2022, January 3). Cómo las tecnologías IoT y IA están transformando la industria minera. Retrieved July 6, 2022, from abcdust: https://abcdust.net/como-las-tecnologias-iot-y-ia-estan-transformando-la-industria-minera/?lang=es.

Adams, T. (2022, February 22). Autonomous mining trucks & haulage systems. Retrieved July 30, 2022, from GRT: https://globalroadtechnology.com/autonomous-mining-trucks-haulage-systems/.

Aerial Insights. (2018, February 11). Cámaras termográficas para drones. Retrieved July 1, 2022, from aerial insights: https://www.aerial-insights.co/blog/camaras-termograficas-para-drones.

Aerial Insights. (2019). ¿Qué es y cómo funciona una cámara multiespectral? Retrieved July 1, 2022, from aerial insights: https://www.aerial-insights.co/blog/camara-multiespectral/.

Ali, D., & Frimpong, S. (2020, April 18). Artificial intelligence, machine learning and process automation: existing knowledge frontier and way forward for mining sector. Artificial Intelligence Review(53), 6025-6042. doi: https://doi.org/10.1007/s10462-020-09841-6.

Aquino, J. A., Fernández, C., & Corona, L. G. (2020). Los gemelos digitales, la última frontera de la ingeniería mecatrónica. X Simposium internacional de investigación multidisciplinaria. Villahermosa. Retrieved June 22, 2022, from https://www.researchgate.net/publication/348850216_Los_gemelos_digitales_la_ultima_frontera_de_la_Ingenieria_Mecatronica/stats#fullTextFileContent.

Byrne, B., & Engdahl, C. (5 de mayo de 2021). Mining 4.0: How innovation is shaping mines of the future. Recuperado el 23 de junio de 2022, de miningdigital.com: https://miningdigital.com/smart-mining/mining-40-how-innovation-shaping-mines-future.

Carvalho, M. (2021, July 15). ¿Cuáles son los impactos de la tecnología en las mineras? [Artículo del blog]. Retrieved June 30, 2021, from https://www.orange-business.com/es/blogs/cuales-son-los-impactos-tecnologia-en-las-mineras.

Castro, F. (2022, January 24). Fotogrametría con drones aplicada a minería. Retrieved July 1, 2022, from linkedin: https://www.linkedin.com/pulse/fotogrametr%C3%ADa-con-drones-aplicada-miner%C3%ADa-frank-castro/?trk=articles_directory&originalSubdomain=es.

Cat. (2018, August 28). Automate drilling with the mining automation leader. Retrieved June 30, 2022, from Cat: https://www.cat.com/en_US/by-industry/mining/surface-mining/surface-technology/command/command-drilling.html.

Cat. (2021, August). Caterpillar expands Cat® Command remote-control operation to excavator line, enhancing safe machine operation. Retrieved June 30, 2022, from Cat: https://www.cat.com/en_US/news/machine-press-releases/caterpillar-expands-cat-command-remote-control-operation-to-excavator-line-enhancing-safe-machine-operation.html.

Caterpillar. (2022, February 3). 500+ autonomous trucks operating worldwide with Cat® Command for hauling. Retrieved June 30, 2022, from Caterpillar: https://www.caterpillar.com/en/news/caterpillarNews/2022/500-autonomous-trucks.html.

CORDIS. (2022, May 25). Real-Time Mineral X-Ray Analysis for Efficient and Sustainable Mining. (Community Research and Development Information Service (CORDIS)) doi: https://doi.org/10.3030/730270.

Cornejo, L. (2022, March 22). Impacto de la digitalización en minerí. Retrieved June 27, 2022, from CAMIPER. Escuela de Altos Estudios: https://camiper.com/investigacion-por-camiper-escuela/impacto-de-la-digitalizacion-en-mineria.

Cornieles, P. (2021, August 2021). Cómo funciona la Visión Artificial o Computer Vision. Retrieved July 8, 2022, from IALATAM: https://ia-latam.com/2021/08/18/como-funciona-la-vision-artificial-o-computer-vision/.

Critchley, L. (2019, July 10). The sensors used in mining applications. Retrieved July 4, 2022, from AzoMining: https://www.azomining.com/Article.aspx?ArticleID=1481

DCC. (2022, February 11). Digital twins to emerge as ESG tool. Retrieved June 28, 2022, from Digging for Climate Change (DCC): https://www.diggingforclimatechange.com/articles/digital-twins-to-emerge-as-esg-tool.

Diniz, A. (2022a, June). Equipamentos autônomos conduzem à transformação digital. (F. E. Alves, Ed.) Brasil mineral, 421, 98-107.

Gartner. (2022, Jun 28). Digital Twin. Retrieved June 29, 2022, from Information Technology Gartner Glossary: https://www.gartner.com/en/information-technology/glossary/digital-twin#:~:text=A%20digital%20twin%20is%20a,organization%2C%20person%20or%20other%20abstraction.

Global Data. (2022, May 22). Global autonomous mining truck population tops 1,000. Retrieved June 30, 2022, from GlobalData: https://www.globaldata.com/data-insights/mining/global-autonomous-mining-truck-population-tops-1000/.

Goundar, S., Bhardwaj, A., Singh, S., & Gururaj, H. L. (2021). Big Data and Big Data Analytics: A Review of Tools and its Application. In S. Goundar, & P. Rayani (Eds.), Applications of Big Data in Large- and Small-Scale System (pp. 1-19). doi: https://doi.org/10.4018/978-1-7998-6673-2.ch001.

Hansford Sensors. (2021, August 19). Sensores de vibración para minería y canteras. Retrieved July 4, 2022, from Hansford Sensors: https://www.hansfordsensors.com/es/industries/mining-and-quarrying-industry/.

Hyder, Z., Siau, K., & Nah, F. (2019). Artificial intelligence, Machine learning, and autonomous technologies in mining industry. Journal of Database Management (JDM), 30(2), 67-79. doi: http://doi.org/10.4018/JDM.2019040104.

IBM. (2019, January 27). Deep learning. Retrieved June 2, 2022, from https://www.ibm.com/co-es/cloud/deep-learning.

IoT Consulting. (2021, March 15). LoRa vs LoRaWAN – ¿Cuál es la diferencia? Retrieved July 4, 2022, from iOTConsulting: https://iotconsulting.tech/lora-vs-lorawan/.

Kiziroglou, M. E., Boyle, D. E., Yeatman, E. M., & Cilliers, J. J. (2017, February). Opportunities for Sensing Systems in Mining. IEEE Transactions on Industrial Informatics, 13(1), 278-286. doi: https://doi.org/10.1109/TII.2016.2636131.

Komat´su. (2021, September 13). Komatsu mining customers surpass 4 billion tons milestone. Retrieved June 30, 2022, from Komat´su: https://www.komatsu.com/en/newsroom/2021/komatsu-mining-customers-surpass-4-billion-tons-milestone/.

Luque, J. (2020). Realidad virtual y realidad aumentada. Revista Digital de ACTA, 19. Retrieved June 22, 2022, from https://www.acta.es/medios/articulos/ciencias_y_tecnologia/063001.pdf.

Martens, E., Prommer, H., Sprocati, R., Sun, J., Dai, X., Rich Crane, R., . . . Fourie, A. (2021, April 30). Toward a more sustainable mining future with electrokinetic in situ leaching. Science Advances, 7(18). doi: https://doi.org/10.1126/sciadv.abf9971.

McGee, P. (2022, February 10). Investors bet on tech revolution to disrupt global mining. Retrieved June 26, 2022, from Financial Times: https://www.ft.com/content/6984328c-fe0f-48f1-a6cd-cd3912b5eac9.

Medina, J. (2012, April 18). ¿Sensor CCD o CMOS? ¿Qué significa todo esto? Retrieved July 4, 2022, from parentesis.com: https://parentesis.com/tutoriales/Sensor_CCD_o_CMOS_Que_significa_todo_esto.

Mining.com. (2022, June 21). Rio Tinto opens Gudai-Darri, its most technologically advanced mine in Pilbara. Retrieved June 21, 2022, from mining.com: https://www.mining.com/rio-tinto-officially-opens-gudai-darri-its-most-technologically-advanced-iron-ore-mine-in-pilbara/?utm_source=Daily_Digest&utm_medium=email&utm_campaign=MNG-DIGESTS&utm_content=rio-tinto-opens-gudaidarri-its-most-technologically-adv.

Mishra, A. (2021, October 12). AI4R2R (AI for Rock to Revenue): A Review of the Applications of AI in Mineral Processing. (L. B. Auret, Ed.) Minerals, 11(10). doi: https://www.mdpi.com/2075-163X/11/10/1118.

Molaei, F., Rahimi, E., Siavoshi, H., Ghaychi, S., & Tenorio, V. (2020, September 3). A Comrehensive Review on Internet of Things (IoT) and its Implications in the Mining Industry. American Journal of Engineering and Applied Sciences, 13(3), 499-515. doi: https://doi.org/10.3844/ajeassp.2020.499.515.

Muñoz, J. (2021, May 7). Voladura 4.0: Digitalización total en el ciclo de producción en mina. Retrieved June 24, 2022, from interempresas.net: https://www.interempresas.net/Mineria/Articulos/350365-Voladura-40-Digitalizacion-total-en-el-ciclo-de-produccion-en-mina.html.

Nad, A., Jooshaki, M., Tuominen, E., Michaux, S., Kirpala, A., & Newcomb, J. (2022, February 7). Digitalization Solutions in the Mineral Processing Industry: The Case of GTK Mintec, Finland. (S. Ogonowski, & D. Foszcz, Eds.) Minerals, 12(2), 21. doi: https://doi.org/10.3390/min12020210.

Next Ore. (2020, July 30). Magnetic resonance technology. Retrieved July 5, 2022, from NextOre: https://nextore.com.au/technology/mag_res_tech/.

Nwaila, G. T., Frimmel, H. E., Zhang, S. E., Bourdeau, J. E., Tolmay, L. C., Durrheim, R. J., & Y., G. (2022, June 24). The minerals industry in the era of digital transition: An energy-efficient and environmentally conscious approach. Resources Policy, 78. doi: https://doi.org/10.1016/j.resourpol.2022.102851.

Okada, K. (2022, May 19). Breakthrough technologies for mineral exploration. (M. Ericsson, & L. Warell, Eds.) Mineral economics. doi: https://doi.org/10.1007/s13563-022-00317-3.

Olivares, P. (2021, June 26). Magnetometría aérea con drones. Retrieved July 1, 2022, from AV3 Aerovisual: https://av3aerovisual.com/magnetometria-aerea-con-drones/.

Parsania, P., & Saradava, K. (2012). Drive-By-Wire systems in automobiles. Journal of systematic computing. 6. Rajkot: VVP Engineering College. Retrieved June 30, 2022, from https://www.researchgate.net/publication/287993938_Drive-By-Wire_Systems_In_Automobiles.

Paulo G. C., Á. R., & Herrero N., J. (2021, August). Preliminary proposal of a model for a holistically harmonious mining industry. Geominas, 49(85), 83-108. Retrieved from www.revistageominas.com.

Pradeepkumar, G., Sanjay, S., Sudharsanaa, N., Suvetha, S., & Ponnusamy, D. (2021). A Smart Helmet for the Mining Industry using LoRaWAN. Journal of Physics: Conference Series. Coimbatore, India: IOP Publishing Ltd. doi: https://doi.org/10.1088/1742-6596/1916/1/012089.

Río Tinto. (2021). Pilbara. Retrieved July 7, 2022, from RioTinto: https://www.riotinto.com/operations/australia/pilbara.

Rose, H., Eldridge, S., & Chapin, L. (2015, October). La internet de las cosas una breve reseña. The Internet Society (ISOC). Retrieved June 22, 2022, from https://www.internetsociety.org/wp-content/uploads/2017/09/report-InternetOfThings-20160817-es-1.pdf.

Saha, S. (2018, December 15). A Comprehensive Guide to Convolutional Neural Networks — the ELI5 way. Retrieved June 2, 2022, from https://towardsdatascience.com/: https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53.

Schodde, R. (2019, October 30). Trends in exploration. Retrieved June 26, 2022, from MinEx Consulting: https://minexconsulting.com/wp-content/uploads/2019/12/IMARC-Presentation-27-Oct-2019-FINAL.pdf.

Schodlok, M. C., Frei, M., & Segl, K. (2022, June 27). Implications of new hyperspectral satellites for raw materials exploration. (M. Ericsson, & L. Warell, Eds.) Mineral Economics. doi: https://doi.org/10.1007/s13563-022-00327-1.

Singh, M., Fuenmayor, E., Hinchy, E., Qiao, Y., Murray, N., & Devine, D. (2021). Digital Twin: Origin to Future. Applied System Innovation, 4(2). doi: https://doi.org/10.3390/asi4020036.

Zamora, M. C. (2017, January/February). La tecnología LiDAR, herramienta útil para el estudio de la biodiversidad. Revista mexicana de ciencias forestales, 8(39), 4-6. Retrieved July 1, 2022, from http://www.scielo.org.mx/scielo.php?script=sci_arttext&pid=S2007-11322017000100004#:~:text=La%20tecnolog%C3%ADa%20LiDAR%20(Light%20Detection,en%20un%20helic%C3%B3ptero%2C%20se%20denomina.

1 Rose, Eldridge & Chapin (2015) explain it as: scenarios in which network connectivity and computing power are extended to objects, sensors, and everyday items not typically thought of as computers, allowing these devices to generate, exchange and consume data with minimal human intervention.

 2 In this regard, Ali & Frimpong (2020) had expressed that "it is a field of computer science that deals with transforming and developing machines capable of performing a certain task without requiring any specific instruction."

3  According to El Naqa & Murphy (2015) cited by Hyder, Siau a& Nah (2019), it is based on computational algorithms that are designed to emulate human intelligence by learning from the surrounding environment using the big data provided to them.

4  According to Ross, et al. (2014) cited by Goundar, Bhardwaj, Singh, & Gururaj (2021) is the process of examining large and varied data sets to uncover hidden patterns, unknown correlations, market trends, customer preferences, and other useful information that can help organizations make more informed business decisions.

5  Luque (2020) explain that it is an environment of scenes or objects with a real appearance, which creates in the user the sensation of being immersed in it. It is a simulated digital reality, in such a way that virtual reality applications immerse the user in an artificial environment, generated by a computer, that simulates reality through the use of interactive devices that send and receive information through the use of sensors and actuators.

6  Luque (2020) has stated that it is the representation of reality, visualized through a technological device, with digital information added by it. In this way, tangible physical elements are combined with virtual elements, thus creating an augmented (enriched) reality in real time. It complements, therefore, the perception of the real world with layers of digital information (still images, sounds, videos, data, 3D models, etc.), which are superimposed on reality (on the perception of the physical world) in real time.

 7 “Concept that arises as a means to represent, through computational technologies, a production system, a robot, a machine, an industrial installation, a new generation production process and that through this it is possible to observe, operate, modify and even carry out with him, a training for the new staff”. (Aquino, Fernandez, & Corona, 2020).

“It is a digital representation of a real-world entity or system. A digital twin implementation is an encapsulated software object or model that reflects a unique physical object, process, organization, person, or other abstraction. Data from multiple digital twins can be aggregated to get a composite view of multiple real-world entities…” (Gartner, 2022).

8  Charge-coupled device. This type of sensor is found in most digital cameras and means that: It is sensitive to light and works as lines of pixels with primary color coverage (RGB); it consumes more power, so the battery may drain faster; it capture a wider range of tones (highlights, midtones, and shadows) in photos. (Medina, 2012).

9  Complementary Metal Oxide Semiconductor. They are found in most current and most professional cameras. They have better performance in a simpler structure, without the need for more equipment, and are less sensitive to light, but much lower power consumption, so the cameras are cheaper. (Medina, 2012).

10  X-ray fluorescence.

11 LoRaWAN is a wide area network within the LPWAN (Low Power Wide Area Network) group. Therefore, the term LoRaWAN refers to a network of LoRa nodes that communicate through gateways and whose messages are managed by a network server. (IoT Consulting, 2021).

12  According to Schodde (2019), 1,070 discoveries have been made in the last decade… but only 19 of these were Level 1 (or World Class); they have also indicated that: “Gold remains the main focus. There seems to be a drop in the size and quality of recent discoveries. It is getting harder and harder to find a Tier 1 deposit. The number of discoveries made each year has been slowly increasing over time.”

13  “Is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image and be able to differentiate one from the other. The pre-processing required in a ConvNet is much lower as compared to other classification algorithms. While in primitive methods filters are hand-engineered, with enough training, ConvNets have the ability to learn these filters/characteristics”. (Saha, 2018).

14  Deep learning is a subset of machine learning (which in turn is part of artificial intelligence) where neural networks learn from large amounts of data. (IBM, 2019).

15 It is currently a subset of ML, more specifically DL. Cornieles (2021) has stated that Computer Vision allows machines to perform image segmentation and pattern recognition, classify objects, track them and detect them. It uses neural networks to sort through massive amounts of data until what you're looking at is understood.

16  Red, Green, Blue. In this case, it refers to the color model of the visible spectrum, which is used in the images that can be obtained with the cameras that drones normally incorporate.

17  Light detection and ranging. “It is a massive remote position measurement system, based on a laser scanning sensor (infrared spectral region) that emits pulses and records the returns against the surface.” (Zamora, 2017).

18  “A multispectral camera, as its name suggests, is a camera that is capable of capturing various spectrums of light.” (Aerial Insights, 2019).

19  They are cameras, which in the case of those that are normal for RPA or UAV can capture the thermal infrared between 8 and 14 microns. (Aerial Insights, 2018).

20  They are sensors that “are designed to detect changes in the Earth's magnetic field. Depending on the type of magnetometer used, the data has to be analyzed and interpreted to obtain the final result: magnetic anomaly maps”. (Olive Groves, 2021).

21  Deutsches Zentrum für Luft- und Raumfahrt e.V. (German Center for Aerospace Travel).