Good morning, my name is Luciana Nedel, I am a professor at the UFRGS Institute of Informatics, and today I'm going to talk about the use of immersive simulators in learning. So we're going to address some facilities, difficulties, and the case study in the planning of battle operations with armored vehicles. So I wanted to start by thanking the invitation of the Southern Military Command, the figure of Colonel Antรดnio de Moura, and the invitation of UFRGS itself, my university, to talk to you in this course.
So I already participated here last year, it was a super interesting experience, I am honored and happy to be here again talking about a slightly different theme than last time. Well, before we start talking about today's class, I just wanted to talk a little bit about my context. So, as I said, I am a professor at the UFRGS Institute of Informatics and I develop research in the VisLab laboratory, where VisLab comes from Visualization, Interaction and Simulation, which are the three topics that we have dealt with in our laboratory.
So we work with information visualization, very frequently with virtual reality visualization, which we call immersive visualization. We work with simulators, with simulation, and we work with interaction. And we understand that there is a very large intersection between these three areas, it is impossible to do visualization and simulation without interaction.
So the areas that we have worked hard on lately are visualization and simulation, both interactive. And then we have in the same laboratory master's students, doctorate students and graduate students who work there. So we're going to start here talking about applied immersive games, let's see a little bit of this concept.
Then I wanted to talk a little bit about the effects of virtual reality on simulators, simulators specifically developed for training, and then I will present the case study in the planning of combined operations, which is an application, the VR Sandbox, which is being developed by a master's student under my guidance, Captain Jerson Neto. So to start, I wanted to introduce this concept of applied games, which we call applied games, which are basically games that have a different purpose from entertainment. So we assume that a game has entertainment as its goal, and the applied games have a different purpose that must be added to the entertainment.
So here we are not just talking about applied games, but about applied immersive games, because they are in virtual reality. So this concept is well known, what we usually call serious games. We wrote a work and published it in 2018 in a magazine, Computer Graphics and Applications, and this work had the goal of reporting a work of biographical research that we did, where we analyzed the use of virtual reality technology in immersive simulators that have a perspective of serious games, of these applied games, as I said.
So basically, we understand that virtual simulators are a safe and low-cost alternative for dangerous task training, or tasks that are carried out in hostile environments, in unsafe environments. We also understand that the use of virtual reality, along with techniques that are used in games, which is the idea of โโgamification, has the potential to improve the user's experience, that is, UX, in virtual environments, due to the fact that it increases realism, increases engagement and increases the motivation of the person who is learning what is being trained. So we studied, from these premises, the impact of virtual reality technology on 46 gamified simulators, that is, that fit into this concept.
And then we classify these works. First, we classify them according to the type of simulation that happened, with immersive simulation. We classify them according to their purpose.
So they could be for training, for learning, for treatment, for self-care, for assessment. We also classified them according to the type of virtual environment being used, what kind of display was used, the level of immersion, the feedback it gave to the user, as well as the ways of interacting. And we also classified them in relation to the profile of the simulator's target audience, if it was a professional, if it was a patient, in the case of simulators for treatment, if they were children, if they were elderly, etc.
And then we made another classification of the same articles, according to the study with the user that had been done, that had been conducted there. So, what kind of analysis was done, was it usability, effectiveness, immersion? How was the design of the experiment made?
Was it experimental or non-experimental? And the participants, what was the profile of the participants and the size of the sample of participants that were used in the tests? Well, then, in our conclusions, I will not go into depth here in this work, but I wanted to show what we discovered.
Basically, 82% of the immersive simulators whose works were published, generated scientific articles, were published in conferences and magazines in the area of โโcomputing. 82% of them are either simulators for training, or for teaching, or for treatment, normally working with the idea of โโrehabilitation. That is, the vast majority of simulators are aimed at these goals.
We learned other things with this work. One of them is that serious immersive games offer results that are similar to the traditionally used methods when it comes to acquiring knowledge from the student. But we also found that they are more effective in retaining knowledge.
People learn in practically the same way as others, but they retain knowledge much more and for longer. We also learned that immersive virtual environments help more than conventional training methods in understanding the situations experienced by users and the processes they are learning. It has some points where it is competitively better than the others.
In conclusion, we also understood that immersive virtual reality improves the results of simulators, increases learning and knowledge retention, improves the results in rehabilitation applications, improves clinical results, but there are some limitations that we need to consider. For example, the issue of motion sickness, which we also call cyber sickness, which is a nausea that some people feel after using an immersive simulator. And we know that access to virtual reality equipment is not yet universal.
They are still restricted. Although year after year, equipment is lowering costs and increasing popularity. Today we have excellent quality head-mounted displays, sold cheaper than a top-of-the-line cell phone.
Anyway, they are not yet widespread in the population, as it would be desirable to work with teaching and training in a generalized way. Well, having learned a little more about simulators, we went to discover other things, that is, what are the effects that virtual reality generates in training simulators. We learned that it is important to understand that it makes sense to work with immersive simulators, but we want to know now exactly where the gains are.
And the results of these works were published in a magazine last year, Computers and Graphics. Our goal here in this work was to explore the perception and acquisition of knowledge. Our goal was to increase the understanding of the effects of virtual reality technology on simulators.
We compared simulators of different levels of fidelity and we carried out two experiments. The first experiment was a user perception experiment. Our goal was to investigate the influence of virtual reality devices on user perception.
To do this, we developed a risk perception evaluation simulator. In the second experiment, we wanted to measure the gain of knowledge. We investigated the effects of four different interaction techniques.
We wanted to understand what effect these interaction techniques had on the user experience, the quality of UX and the gain of knowledge. To do this, we implemented a lightning switch simulator, which is a procedure that is not known to most people. Our first experiment was a perception experiment.
It considered three different levels of immersion. In the immersive version, people used the head-mounted display. In the semi-immersive version, they used a display wall.
In the non-immersive version, they used a desktop monitor. We had two hypotheses here. The first is that devices that provide greater immersion improve user performance, that is, the more immersed, the better the performance.
We thought that cyber-sickness would not impact user performance, believing that the equipment available today is of good quality enough for most people not to suffer from nausea. We conducted the test with 61 participants. Most of them were lay participants, they were not our target audience.
In the age range of 19 to 63, it was a very large sample. We measured a number of things. We measured cyber-sickness using a questionnaire suitable for this.
We measured the sense of presence, the sense of immersion. We collected information about the quality of the user's experience. We measured how long each user took and the number of mistakes they made.
In relation to our hypotheses, the first hypothesis, which said that devices that offer greater immersion improve user performance, we were unable to prove. Basically, the experiment took 5 to 15 minutes, which was a relatively short duration. We were unable to see a difference in performance in relation to this.
The second hypothesis was that cyber-sickness would not impact user performance. And it really did not impact. People were able to conclude the experiment without any side effects.
We were unable to see a correlation in relation to time and performance. We also came to other important qualitative conclusions. Users were satisfied with the experience.
It was a positive experience for everyone. They had a greater sense of presence and immersion with the head-mounted display. That is, in the immersive condition, they felt more present and immersed.
And the feeling of being present in that world increased. These were perceptions about the experiment that were very positive. In experiment 2, we wanted to measure the ability to acquire knowledge through the use of immersive virtual reality.
In this case, all conditions were immersive, all with the head-mounted display. And we had four different ways of interacting. In Condition A, we worked with a pointer light, where the user had to point to objects.
And in the virtual world, we used a technique called "walk in place", where the person walks naturally, but in the same place. In Condition B, we used the same requesting technique, but moved using a joystick. In Condition C, we used the direction of the head, where the user looks and selects objects.
And we used the walking technique in the same place to move. And in the last Condition, Condition D, we used the orientation of the body. The choice of objects was made using the movement of the head, but the locomotion happened using a joystick.
We established three hypotheses to be verified. The first is that the more natural the movement, the better the performance in the task. The second is that the use of devices that require the movement of hands and feet will not result in cyber-sickness.
We always believed that cyber-sickness was not important here. And the third hypothesis is that techniques that induce a higher cognitive load will impair user performance. In other words, if users need to divide their attention between doing the task they should do in the simulator, which is learning, and how they manipulate interfaces, we understood that this would generate a problem.
Throughout the experiment, we measured cyber-sickness, cognitive overload, engagement, acquisition of knowledge. We did two tests, one right after learning to use the simulator, and another two weeks later. We also measured the time it took people to do the tasks and the accuracy.
For this, we did tests with 46 participants between the ages of 18 and 33. The results of this experiment did not achieve statistical significance for any of our hypotheses. We have reasons for this, we will talk a little about them now.
But that does not mean that the results were negative. Our tests lasted 20 to 33 minutes. The D condition was the slowest.
Remembering that the D condition was the one that used the joystick, used the head movement to select, to point to things. So it was a very unnatural condition and ended up resulting in the worst performance. There was 95% accuracy in the tasks.
People were able to do the tasks in any of the four conditions that were tested. More than 50% of the users were able to acquire the knowledge. There were no significant differences between the conditions.
In the evaluation of the interfaces, there was a score of 49. 2 to 62. 4 in the SUS test.
Which is a very positive result, since the B condition, which was a very unnatural condition, had a very good score. Contrary to what we imagined. So there are several reasons for not being able to demonstrate the hypotheses from a statistical point of view.
One of them is that the task was not so difficult. People were able to learn the task in any of the conditions. Or our sample was relatively small.
There were 46 people, so a little more than 10 people for each of the conditions is really small as a sample. What did we learn in this test process? The first learning is that disorientation is probably a consequence of the user's ignorance of their surroundings.
In other words, we need to increase knowledge to improve the user's experience. When the user is in a virtual environment that reports to a real environment that he knows, this ends up having a positive impact on the result. The second learning is that the use of the most natural possible interaction techniques continues to be the best choice.
So we had a qualitative feedback on this, despite not being able to demonstrate quantitatively from a statistical point of view. The third learning is that perception is more affected by the user's experience on the simulator topic than by the technology that is being used. In other words, familiarity with virtual objects improves the user's experience.
Again, both our first and third lessons lead us to believe that applying the tests to people who are the target audience of the simulator will bring us a better result, since they are people who are already more familiar with the environment, objects and tasks. And finally, the fourth lesson learned is that gaming experience plays an important role in virtual reality simulation. The familiarity of users with video games can reduce the issue of cybersickness and adaptation to this new platform.
So, concluding this part of the work here, we carried out two studies with the user. In total, we did tests with 107 test subjects. The assessment of the effects of three different display devices was done in the first experiment to measure perceptive learning.
In the second experiment, we evaluated the effects of four different interaction and locomotion techniques. In this case, the study was for perception and retention of knowledge. We learned several lessons, as we just reported, about the choice of one configuration at the expense of the other to develop simulators.
And we recognize some limitations in our work that lead us to possible future work. First, as it is about learning, I could not work with a user assessment considering a sample within subjects. Each user did the simulation once, and as the goal was learning, he learned.
After that, I could no longer submit him to the same task as another simulator condition, because the learning had already happened. This made our samples smaller and made it difficult for us to find statistical significance in the tests. The participants we used here were not the target users.
We were not able to test with people who were being trained for that goal, which was to learn how to change a lightning rod, for example. This may also have had an effect here. So we see it as a future work.
We would like to replicate these experiments with larger groups and groups formed by the target population. The results we measured are always very dependent on the devices that were used. The devices evolve, and with that we need to do new tests.
At this stage of maturity, we still need to do tests. I can't say head-mounted display gives a better effect than a display wall. I need to do new tests every time.
To conclude, I would like to make some comments. First, the use of immersive simulators is increasingly seen as a promising possibility for training, bringing clear benefits. We have seen this, we continue to work in this area and have strongly perceived it.
Our results also indicate a promising future. As I mentioned, new devices offer new possibilities for interaction, visualization, and the need for better interaction techniques associated with them. With this, new tests must be done as the devices evolve, and they have evolved very quickly.
We are now in the third comment, where everyone is talking a lot about the metaverse as a reality. The metaverse is not yet a reality, it is a vision of the future. But we believe that the metaverse will facilitate integration between simulators and users.
They will go from one simulator to another with greater ease, as soon as the metaverse concept is realized. We still have little knowledge about the prolonged use effect of virtual reality on users. Most immersive virtual reality experiences last up to 45 minutes.
15 to 45 minutes in general, some last 1 hour, 1 hour and 15 minutes, but not more than that. We will need to learn a little more about this. Can I use a simulator in virtual reality?
How long do people last? What is the ideal time to use it? And finally, we also understood, with these initial experiments, that the development of simulators must involve the target users from their conception.
We are currently developing two lines of simulators. One of them in simulators in the medical area, the other in simulators in the military area. In these new projects, we have involved our participants, the expert in the area, the target user in general, throughout the process, from the conception to the preliminary tests, then moving on to another type of test.
Now, to illustrate this work that we have been doing lately, working close to the target users, I will now give the floor to Captain Jerson Neto, who is the captain in Agulhas Negras Military Academy, and he is responsible for training cadets. Jerson is my master's student, he has been with us here at UFRGS since November, December of last year, doing a super interesting job, and I will let him speak now. Before I conclude my part, I want to thank the people who were involved in this part of the work here, Aline Menin, who was my master's student, then my doctorate student, Rafael Torchelsky, who is my colleague, professor at UFPEL, I would like to thank the colleagues from our lab, VISLAB, as well as the 107 volunteers who participated in the tests.
So, I would like to thank you here, I will now give the floor to Jerson, and then we will both be available here to answer any questions, any curiosities. Thank you very much. First of all, good morning everyone, I am Captain of Cavalry Jerson Geraldo Neto, Master in Computing at UFRGS, and I am currently studying at the School of Official Improvement (ESAO) here in Rio de Janeiro.
In the next slides, I will be sharing with you a little bit of the work that I have developed over these two years of research at UFRGS, referring basically to how virtual reality can become a tool of assistance to individuals who have different levels of spatial habilities within the context of strategic planning. First, I will start with a brief introduction. So, first of all, we have to take into account that planning at all levels, so to speak, whether in the military context or not, it is based on a data analysis, before the decision-making itself.
In the military context, it basically affects two primary elements. When we enter the field, mainly of the elements of a lower scale, such as the platoon level, for example, where the interaction between the relief, in this case the terrain, and the field of view of the armaments, or even of the armor, is preponderant. Having in mind this interaction, we have to study these elements to make the best possible decision.
And this has to be done based on the analysis of information that we find mainly on maps, so to speak. And this interpretation directly affects the result. So, this interaction between both is essential.
And how can we get the best information within this context? Today we basically use two elements in this planning, so to speak, which are the most usual. I would cite the topographic maps, the upper image of this slide, which shows the level curves, vegetation, as well as urbanization and hydrographic elements, and high-resolution satellite images, which also help in this process of choosing positions.
However, this interpretation of maps, which has been the result of studies since 2005 by Ishikawa, does not show that it is not a very easy task, and it is directly associated with the spatial ability of each one. That is, some will have more ease, others will have more difficulty in this process of analysis of this type of information, which can obviously harm the planning in itself. So that's why we resort to 3D reconstruction.
Why? Interpreting 2D and mentally representing it in a three-dimensional way is not a simple task. So manual reconstruction, as it is done from what we call a sandbox, is something that aims to reduce the negative impact of this interpretation and reduce these errors.
However, because it is a manual task, it remains a problem. In this context, can technology not be a solution? So the goal of the research that we developed over these two years was basically to develop something from virtual reality that would be able to replace this process, this process of manual land reconstruction production.
And what were the main objectives? They were to investigate the influence of virtual reality in the face of the conventional method that we usually use in this planning process, and how to seek to help these users who have different levels of spatial ability in this planning process, verifying if there is a difference in the result of these plannings, in virtue of these levels of ability, and obviously seek, through virtual reality, to give them the ability to improve their decision-making process. For this, some decisions were made in the development of the project.
So, in general, it was an interface where the main objective was to visualize the interaction of data, of range and terrain, with the use of digital models, elaborated from real information, to offer greater credibility and reliability to the process. And the interpretation of maps and spatial abilities basically relied on what? In the forms of manipulation and visualization, so that we had something as close to the real as possible, compared to the conventional processes, and what was actually developed in virtual reality, to this problem of different scales and distance perception that occurs in the virtual world, and also occurs in the real world, but in order to reduce these negative impacts, and seeking to adapt to these individual characteristics, so that everyone was able to use this tool and gain the maximum possible performance.
So, this research basically started in mid-January 2023, with an initial application, whose goal was basically to study if there was really something that we could take advantage of in virtual reality, aiming at this interaction of range and terrain that was useful. So, an experiment was carried out, a prototype test, which was even the target of presentation in the last edition of our extension course. And after this first experiment, which went very well, we started some adjustments, updates, seeking to investigate if, in some way, this developed application could help those individuals who have difficulty in this decision-making process.
So, new tool updates were developed and a new experiment was carried out, with the goal of verifying how impactful virtual reality would be in this strategic planning, obviously taking into account the different levels of spatial abilities of individuals. So, what is this tool? Basically, we call it VR Sandbox, as the name itself says, it is a virtual reality sandbox, with the goal, basically, of bringing this visualization of range and terrain data in its interaction, based on three, so to speak, fundamental pillars, which would be the digital terrain, a database associated with 3D models, and the interaction, always seeking to have simplicity and using real data to offer accuracy and reliability to the process.
So, the digital terrain, it was made from the use of geo-referenced data obtained by satellites, an elevation digital model that, through the engine used for the development of the application, where we were able to rebuild this terrain in a more reliable way, the use of topographic maps, which is already something usually used in conventional planning, as well as associated with high-resolution satellite images. As we can see in the image on the right, the terrain modeled from this elevation digital model, with these textures, so to speak, either from the topographic map or from the high-resolution satellite image. So, with this combination of elements, we will be able to offer greater accuracy and realism to the production of this digital terrain.
And the 3D and Dataset models, we use realistic virtual models on a real scale, which were obviously manipulated in order to make them easier to manipulate during the virtual scene, as is done in the conventional process, where miniatures are positioned on the relief, but with some updating of the scale in order to leave it in the natural size in relation to the relief. However, we could not visualize this reach. So, for this, it was necessary to build a database, with some elevation elements, field of view, maximum reach of utilization, as well as the dimensions of this armored vehicle itself, so to speak.
And as a way to visualize this information, we chose a light cone, a spotlight, so that these elements that we set up this database were able to be visualized in the virtual scene as a light. So, when the element was positioned on the relief, as we can see in the lower image of this slide, a red light, we see the points where it is possible to observe, in that chosen position, and where it is not possible, obviously, due to the blind spots, places where the light is not being illuminated, referring to the reach. And the interaction, we seek to base as much as possible on what is the conventional air box, the use of a reduced scale terrain, the manipulation of the miniatures, as well as some auxiliary planning elements, such as lines to represent the coordinates of latitude and longitude, to assist in navigation, some drawing tool to enable the drawing of paths, objectives, which are elements associated with military planning.
And we seek to offer the maximum of naturalness, freedom of movement and manipulation, so that the user has the freedom to walk around the scene, to take the objects using the controls in his or her hands and put them on the relief, among others. After all this initial development, as has already been said in the other, there was the prototype test, it was verified that it was useful, then we went to the experiment 2, which had the purpose of investigating the influence of spatial ability in planning, verifying the impact of immersive reality in the development of capacities, and seeking to assist individuals with difficulties in this process. First of all, it was necessary for us to understand how our audience would behave, that is, what is the level of spatial ability and interpretation of maps of that universe that we intended to select.
For us to understand what would be the greatest results, what would be the averages, what would be the lowest, so that we have the ability to select those individuals and test our hypotheses based on that. Initially, we released two main tests. The spatial ability test, which is the cross-section, as we can see in the image on the right, a sectioned image, which would be a representation in 3D of some element, of a cube cut with a cylinder, and a sectioned view of that fraction, and some options to select.
And also the topographic maps assessment, which is a map interpretation test, as we have said, with a series of tests, and one of them was this, to obviously imagine being in one of the positions there, observing a certain level curve meeting, how would the visualization of this terrain be represented in the real world. So these tests were widely disseminated and a total of 106 participants were volunteers, basically cadets from our military academy and some instructors as well. And we understood from these results, from these users, what were the levels of spatial ability of our individuals.
So we already knew how we could expect the highest results, the averages, as well as the lows. Dividing there in quarters, based on these results, we see three large groups of concentration. The highest varying between 34 and 42 hits, the averages between 25 and 33, and the lowest between 9 and 24.
Now understanding how our audience behaves in this spatial ability test and map interpretation, we were able to select those who could, let's say, participate in our experiment, in order to meet the conditions. So first, three large groups based on the levels of spatial abilities, A, B and C, each of them receiving an identification, the IDs varying from 1 to 12 within each group. So A from 1 to 12, B from 13 to 24 and C from 25 to 36.
And starting the experiment, either by virtual reality or by the conventional process and having this variation. And to avoid contamination of results, obviously, we developed two different scenes. With the same difficulty, with the same purpose, but with different maps, different regions, so to speak.
And to meet all these conditions, we needed to select 36 users of those 106 who participated in that initial test. And what were the data collected in this experiment? So basically, the positions chosen in both processes, normalized and compared to each other, to verify if the users of these different groups of abilities tended to choose different positions, and if there was any difference between them in this process of choosing positions.
The evaluations, which would basically be another individual, an instructor who was invited, to evaluate the positions chosen by the users in the different methods, without knowing which one they were dealing with, based on technical criteria, quantifying these spans in a scale of 0 to 10. We wanted to know if, either in virtual reality or in the conventional process, these individuals with different levels of spatial abilities, if their planning result was different too. If there was a difference and a correlation between spatial ability and strategic planning.
So that we could verify if the user who has a lower spatial ability level, he has more difficulty, more ease or indifference, in the case of choosing the positions in the planning. And as well as some impressions of users from usability. Collecting impressions of users about the use of the tool, its utility, as well as among other aspects.
Regarding positioning, we already verified that there was a difference. This here on map A, later we will possibly show map B. We can see that the little balls represent the individuals with the highest levels, the blue balls, the average ones, the red ones, and the lowest ones, the green ones.
We see that there was a distinction between the positions chosen by the user in both processes. In the conventional, there was an even greater distinction, especially in the green balls, if we observe, with more distinct positions in relation to the red and blue ones. And when we vary to virtual reality, there is still a difference in these choices, but they are, in a way, closer to each other.
This, obviously, without considering the issues related to technical criteria. Here, only the positions themselves. And when we vary to map B, we check that these distinctions are still maintained between the different groups in both processes.
And we see the trend of the positions still grouping in virtual reality in relation to what was done in the conventional. And when we put, without considering the levels and abilities, only the conventional process or virtual reality in both maps, we see that these distinctions are even more noticeable. Each map had a different trend, so if we check map A, the trend was, in the conventional process, the points being more grouped and in virtual reality more distant, in map B the process was the opposite.
Remembering that this does not mean that the positions are better or worse, considering that there are technical criteria to assess this choice of positions. Then, when we go to the evaluations, we check that there was also a big difference in the grades in both processes. So, if we check there, the grades in the conventional process had a much greater variation in relation to the virtual reality process, and the virtual reality had higher averages in relation to the conventional.
This without considering the levels of spatial abilities that I will show later. Here we can see how these grades were concentrated within the different ability groups, and what were the trends of these grades, as well as the significant differences between them. So, we check, starting on the right, by the conventional grades, that the grades have a great variation, so there is a distinction between the groups, that is, there is a correlation between spatial ability and map interpretation, and the planning process of choosing positions.
And the individuals of the groups, they have a significant difference between them, especially the individuals of the lowest and highest levels, with a significant difference, with the statistical test that was applied in this case, which was the Tuck test, in relation to the difference in the averages of the groups, we verified that the average was greater than 0. 001. And, in this case, there was also a difference between groups A and B, which also had a significant difference.
However, when we move to virtual reality, we see that these grades are getting closer, and there is only one significant difference that still remained, despite everything, between group A, the highest, and the lowest. But between the highest group and the average, there was no significant difference, which shows, in a way, that there was an improvement in ability due to the use of virtual reality. And there was also the presence of some positive outliers, so to speak.
They are individuals who, despite belonging to groups B and C, had a performance far above their average, which are those highlights, those black dots on the graph on the left. So it means that these individuals had a result far above the others, showing that, in a way, virtual reality destroyed a capacity that they did not have before. And in relation to usability, what were the user's perceptions during the use of the tool?
Everyone, in a way, said that it was a useful and easy-to-use application, compared to the conventional process, that their vision of this process was improved, either by visualizing the issue of the interaction of the relief with the reach, or even of the relief, probably, being as trustworthy as possible to the real terrain, that there was an economy of mental capacity and time in relation to conventional methods. Obviously, the time spent on the conventional process is large, it is considered high, considering either the modeling of the terrain in a manual way, or the study of the map to understand which were the best positions. That there was a significant change in the planning, this we can see in the choices of the two processes, in the raw choices, so to speak, that there was a variation between what was chosen in the conventional process and in virtual reality, and that the richness of details made the planning, in a way, safer.
Finally, what was the general conclusion of this work? That this conception of this application in virtual reality, from the use of real data, the identification of relief, reach, interaction, showed us the impact of immersion in this process, of how it can be beneficial for planning and also, in a way, for teaching. And the results obtained, as they say, that show us that there is a correlation between spatial skills and strategic planning, which can be something to observe, in the case of understanding if I need, in a certain way, to study better how to improve the individual's ability, if I apply this spatial skills test, I already know what will probably be the result of his planning, and I already know how I can help him, in a way.
And as we saw the impact of virtual reality in decision making, we can verify that it can also be an aid tool for those individuals who have difficulties, to better understand the behavior of this interaction of relief and reach. And, in a way, it was shown that virtual reality reduces these barriers between the different levels of skills, which is also very beneficial. And what are the future works that can still be done to further improve the use of virtual reality or study even better what can still be done?
So, in relation to the experiment, we can increase the size of the sample, looking for more evidence of the impact of virtual reality, as well as its benefits. It is always good to highlight this. Refinement adjustments in the interface, which had problems, obviously, with respect to the aspects of interaction in some moments.
So, this is always welcome. It comes with the implementation of new capabilities, so to speak. And following the use of the tool as a teaching tool may be something that can be quite useful over the years.
In a way, we can verify if, in fact, virtual reality can improve the skills of those individuals who have difficulties, especially when it comes to a military academy, for example, where we have individuals who will have more difficulty in this planning process and others who will have more ease. And since the objective of the military academy is to train the commanders of small factions and that everyone knows how to interpret the terrain in the best possible way, maybe this technology can be quite useful in this process. I thank everyone's attention.
These are my contacts for those who want to interact with me, be it my functional email or even my email here at the university. Thank you very much for your attention. I am available for any questions.