Entering the Augmented Era: Immersive and Interactive Virtual Reality for Battery Education and Research

We present a series of innovative serious games we develop since four years using Virtual Reality (VR) technology to teach battery concepts at the University (from undergraduate to doctorate levels) and also to the general public in the context of science festivals and other events. These serious games allow interacting with battery materials, electrodes and cells in an immersive way. They allow experiencing impossible situations in real life, such as building with hands battery active material crystal structures at the nanometer scale, flying inside battery composite electrodes to calculate their geometrical tortuosities at the micrometer scale, experiencing the electrochemical behavior of different battery types by driving an electric vehicle and interacting with a virtual smart electrical grid impacted by 3D-printed devices operated from the real world. Such serious games embed mathematical models with different levels of complexity representing the physical processes at different scales. We describe the technical characteristics of our VR serious games and their teaching goals, and we provide some discussion about their impact on the motivation, engagement and learning following four years of experimentation with them. Finally, we discuss why our VR serious games have also the potential to pave the way towards an augmented era in the battery field by supporting the R&D activities carried out by scientists and engineers. 1. Energy storage and virtual reality Energy storage is one of the most prominent challenges Humanity has to face in the 21 century. This is mainly due to the increasing use of smart grids and of intermittent renewable energies driven by the limitation of fossil fuels and the climate change. Electrochemical energy storage technologies such as 2 lithium-based rechargeable batteries are called to play a major role to address this challenge, due to the simplicity of their overall operation principles and high energy densities. Since their first commercialization in 1991 by Sony, Lithium Ion Batteries (LIBs) have triggered the emergence of a wide spectrum of portable devices and they are nowadays the hobbyhorse of the renaissance of the Electric Vehicles (EVs). The desired massive electrification of the transportation sector still requires better rechargeable battery technologies than current LIBs in terms of specific energy, cost, recyclability and safety. To make this happen, the design of advanced LIBs with a new generation of electrode materials and smart functionalities (e.g. embedded state of health sensors and self-healing actuators) is required. Other lithium-based rechargeable battery technologies have emerged at the laboratory scale and have been the focus of intense studies in the last decades such as Lithium Sulfur Batteries (LSBs) and Lithium Oxygen Batteries (LOBs) due to their higher theoretical specific energies than the ones achievable in current LIBs. However, significant technical challenges still persist in LSBs and LOBs, such as the premature capacity fading in the former and the severe instability of the electrolyte and poor rechargeability in the latter. In general, any scientific approach adopted to try to overcome the aforementioned technical challenges as well as to design and optimize any kind of electrochemical energy storage device, requires transdisciplinary efforts, encompassing at least materials science, chemistry, physics and engineering. Indeed, lithiumbased batteries are made of multiple materials and their operation implicate numerous physicochemical mechanisms occurring simultaneously at multiple spatial scales. To successfully implement the required transdisciplinary approaches and to favor the invention of disruptive energy storage technologies, it is of paramount importance to encourage the emergence of tools that can ease the inspiration by the current and the future generations of battery scientists. Virtual Reality (VR) technology constitutes one of such tools because it can ease training, education and stimulate creativity in research. VR environments were adopted very early for training in abstract or complex situations. They have been revealed very useful for training in the cases when the possibility to perform in vivo situation is either difficult, expensive or dangerous (e.g. in spatial, military and nuclear areas). The immersive experience in an environment inspired by physically realistic phenomena in which it is possible to interact represents an important asset of VR: this helps users at developing an intuitive knowledge about the simulated system. Thanks to the recent technical progresses, VR hardware has become more compact and accessible to much wider audiences: Oculus and HTC Vive are examples of commercial VR hardware. These technologies have head-mounted displays that allow easy interaction with the virtual environment by using controllers equipped with motion tracking (Figure 1). Figure 1. a) Illustration of the HTC Vive VR hardware (adapted from Ref. [20]); b) overall principles (need of a head-mounted display, virtual environment in 360°, full immersion, full interaction using controllers); c) typical location of the base stations and the user (adapted from Ref. [20]); d) a VR user in action.


Abstract:
We present a series of innovative serious games we develop since four years using Virtual Reality (VR) technology to teach battery concepts at the University (from undergraduate to doctorate levels) and also to the general public in the context of science festivals and other events. These serious games allow interacting with battery materials, electrodes and cells in an immersive way. They allow experiencing impossible situations in real life, such as building with hands battery active material crystal structures at the nanometer scale, flying inside battery composite electrodes to calculate their geometrical tortuosities at the micrometer scale, experiencing the electrochemical behavior of different battery types by driving an electric vehicle and interacting with a virtual smart electrical grid impacted by 3D-printed devices operated from the real world. Such serious games embed mathematical models with different levels of complexity representing the physical processes at different scales. We describe the technical characteristics of our VR serious games and their teaching goals, and we provide some discussion about their impact on the motivation, engagement and learning following four years of experimentation with them. Finally, we discuss why our VR serious games have also the potential to pave the way towards an augmented era in the battery field by supporting the R&D activities carried out by scientists and engineers.

Energy storage and virtual reality
Energy storage is one of the most prominent challenges Humanity has to face in the 21 st century. [1] This is mainly due to the increasing use of smart grids and of intermittent renewable energies driven by the limitation of fossil fuels and the climate change. [2] Electrochemical energy storage technologies such as lithium-based rechargeable batteries are called to play a major role to address this challenge, due to the simplicity of their overall operation principles and high energy densities. Since their first commercialization in 1991 by Sony, Lithium Ion Batteries (LIBs) have triggered the emergence of a wide spectrum of portable devices and they are nowadays the hobbyhorse of the renaissance of the Electric Vehicles (EVs). [3][4][5] The desired massive electrification of the transportation sector still requires better rechargeable battery technologies than current LIBs in terms of specific energy, cost, recyclability and safety. To make this happen, the design of advanced LIBs with a new generation of electrode materials and smart functionalities (e.g. embedded state of health sensors and self-healing actuators) is required. [6] Other lithium-based rechargeable battery technologies have emerged at the laboratory scale and have been the focus of intense studies in the last decades such as Lithium Sulfur Batteries (LSBs) and Lithium Oxygen Batteries (LOBs) due to their higher theoretical specific energies than the ones achievable in current LIBs. [7,8] However, significant technical challenges still persist in LSBs and LOBs, such as the premature capacity fading in the former and the severe instability of the electrolyte and poor rechargeability in the latter. [9,10] In general, any scientific approach adopted to try to overcome the aforementioned technical challenges as well as to design and optimize any kind of electrochemical energy storage device, requires transdisciplinary efforts, encompassing at least materials science, chemistry, physics and engineering. Indeed, lithiumbased batteries are made of multiple materials and their operation implicate numerous physicochemical mechanisms occurring simultaneously at multiple spatial scales. [11] To successfully implement the required transdisciplinary approaches and to favor the invention of disruptive energy storage technologies, it is of paramount importance to encourage the emergence of tools that can ease the inspiration by the current and the future generations of battery scientists. Virtual Reality (VR) technology constitutes one of such tools because it can ease training, education and stimulate creativity in research.
VR environments were adopted very early for training in abstract or complex situations. [12] They have been revealed very useful for training in the cases when the possibility to perform in vivo situation is either difficult, expensive or dangerous (e.g. in spatial, military and nuclear areas). [13][14][15][16][17] The immersive experience in an environment inspired by physically realistic phenomena in which it is possible to interact represents an important asset of VR: this helps users at developing an intuitive knowledge about the simulated system. [18] Thanks to the recent technical progresses, VR hardware has become more compact and accessible to much wider audiences: Oculus and HTC Vive are examples of commercial VR hardware. [19][20] These technologies have head-mounted displays that allow easy interaction with the virtual environment by using controllers equipped with motion tracking (Figure 1). In the recent years, VR technology started to be significantly used for educational and research purposes in hard sciences such as mathematics, physics and chemistry. [21][22][23][24][25] In chemistry, especially, researchers have proposed VR tools to visualize quantum mechanics and molecular dynamics simulation results or to teach experiments in immersive and interactive way. [26][27][28][29][30][31][32] In the battery field, we can think that VR has also strong potential to provide a fully immersive and interactive experience that can tremendously ease the understanding of concepts behind materials, components, cell and packs working principles.
VR could be used to put the users in situations that are impossible in reality, such as manipulating and navigating inside a material of few micrometer size by interacting and by measuring the consequences of the interactions in real time. Surprisingly, despite these tremendous promises, the use of VR in the battery field has never been reported. Instead, the most used traditional way providing virtual experience in the battery field is computational modeling. Since more than 50 years computational models have revealed to be useful simulation tools to understand battery operation principles and to carry out their design and optimization. [11,33] Such models are based on mathematical equations describing physicochemical mechanisms which are solved by using informatics programs. Therefore, the models constitute virtual materials or batteries with which one can perform virtual experiments, such as investigating how operation conditions (e.g. applied current density) impact outcomes such as the capacity or the aging. Numerous models have been proposed under several geometric representations of the cells and the electrodes within, in one dimension (1D), two dimensions (2D) or three dimensions (3D). [11] However, the efficient development and use of these models require some programming skills. Moreover, the visualization of results arising from 3D models (e.g. spatial distribution of chemical species in the electrodes) is not trivial: images remain "confined" in 2D computer screens. This issue of 2D representation of 3D objects make that often students and battery researchers are confronted to difficulties in the abstraction and conceptualization in three dimensions of battery materials and components, such as the active material crystal structures at the nanoscale, the composite electrode porosity at the mesoscale and the theoretical operation principles of these batteries in real applications. Some educators and science communicators have used classical 3D glasses (like the ones to watch movies in theaters), [34] but these glasses do not permit individual or collective interaction with the objects and immersion into them.
In this Concept, we report six VR serious games we started to develop four years ago, allowing students and researchers to interact in an immersive and realistic way with virtual battery materials, composite electrodes, battery cells in EVs and smart electrical grids, as well as with the mathematical models used in the VR environment simulations. With serious games [35][36] we refer here to games designed for learning the operation principles of rechargeable batteries, with the aim at rising the motivation, the engagement and the learning efficiency by students and by the general public, and also to stimulate the creativity of battery researchers. Our pedagogy activities using these VR serious games were recently recognized with one of the French National Prizes for Pedagogy Innovation in 2019 (PEPS 2019). [37][38][39][40] The goal of this Concept is to present the main characteristics and pedagogical goals of these VR serious games, as well as some illustrations of their utilization in the last four years. Some discussion about the impact on the motivation and learning by students and other users in general is also provided, without the intention to share here detailed ergonomic and psychological assessments in view of the readership of this Journal. Such assessments are being shared by us elsewhere (see for instance Ref. [41]). We also discuss the potentialities of these VR tools to boost battery R&D.

Battery virtual reality serious games
All our VR serious games were coded using Unity programming language and Revia ® technology. [   The Crystal VR serious game, developed by us in 2019, addresses the structural (crystallographic) properties of the active materials. [43] Teaching crystallography to students often faces a dilemma. How to draw in two dimensions, symmetry operations and crystal structures which are by nature three-dimensional? Of course, perspective drawings are used, but still, some students and even battery researchers are not at all able to clearly see (and thus to understand) a 3D representation on a 2D plan.
The crystal structure of any material is the periodic repeatability of a so-called "unit cell" in the three directions of space. This "unit cell" is constituted of atoms (same or different chemical nature) related by symmetry operations. If we remove the elements of symmetry, we obtain what is called the "asymmetric unit". Taking the example of table salt (NaCl), we can see that the "asymmetric unit" is made from only two atoms (one sodium atom and one chlorine atom depicted as yellow and green spheres respectively in Figure 3). However, adding symmetry operations, it turns out that the chlorine is actually surrounded by six sodium atoms in a very specific arrangement called octahedron in the "unit cell".
Finally, all those "unit cells" are repeated in the three directions of space, forming in fine the salt crystal.
In crystallography, symmetry operations are divided into four main categories: • Mirrors (labelled "m");  As mentioned earlier, explaining and visualizing the symmetry operations is not an easy task. As it can be seen on Figure 4, being able to differentiate a mirror and a rotation axis of order 2 (2π/2=180°) is not trivial, especially when using spheres (which are highly symmetrical objects). Another feature allows to change the chemical nature of the different atoms and also to build its own crystal structure from an empty unit cell.
Both Crystal workshops are very versatile so that changes and improvement can be easily made. The Crystal Structure workshop has been developed as a tool to better visualize crystal structures. We can imagine creating exercises in which students will be asked to construct specific crystal structure types. They should select the proper unit cell, the proper atoms and element of symmetry in order to build the proper crystal structure. Another improvement could be the display of the X-ray diffraction pattern created from the selected (or created crystal structure). X-ray diffraction is used every day in laboratories working in the battery field. An X-ray source illuminates a sample, which reflects (diffracts) the X-ray beam. We obtain the so-called X-ray diffraction pattern consisting of a succession of peaks with different intensities, from which crystallographers can deduce the crystal structure of given material. The position of the peaks are related to the size of the "unit cell" described above and their intensities to the chemical nature and the position of the atoms into the "unit cell". But, it is also possible, knowing the exact crystal structure, to predict and calculate what the X-ray diffraction pattern should looks like. It will be indeed very interesting for students to be able to see such simulated patterns corresponding to the crystal structure they are looking at and observe how the pattern evolves when they change the chemical nature of the atoms, their positions and so on.
Furthermore, battery researchers are really interested in being able to "fly" into crystals. It is possible to imagine an evolution of Crystal VR to see and follow the diffusion path that a Lithium or a Sodium ion is taking when migrating through the active material.
This could allow researchers to better understand why a given electrode material is more efficient than another one and consequently, could help them to design new materials with enhanced properties.
The effective diffusion coefficient of the lithium ions in the electrolyte is given by [46] (2) where the porosity ε is defined as the ratio between the volume occupied with electrolyte in the electrode and the volume of the electrode.
Even if the geometrical tortuosity concept is easy to understand in two dimensions (Figure 7b), its understanding in three dimensions is not trivial at all as it requires imagining threedimensional paths for no discontinuous transport between one side and the other one of the electrodes. Numerous students and battery researchers have difficulties at imagining and visualizing such three-dimensional paths. This is why numerous theories exist aiming at calculating the geometrical tortuosity of electrodes as function of their porosity, the most famous one being the Bruggeman relation postulating that [47] √ (3)  (1)) are also indicated.
Lelectrode stands for the electrode thickness; b) traditional (bi-dimensional) way of illustrating the concept of "tortuous path" (arrow between points A and B). Note that the porous channel on the top has infinity tortuosity whereas the one at the bottom has lower tortuosity than the one in the middle.
However, these theories assume the electrodes to have unrealistic ideal geometries. In the case of the Bruggeman relation above, it holds when the electrode is constituted by spherical particles and with low volume fraction. It has been demonstrated that other particle shapes lead to other mathematical expressions relating the geometrical tortuosity to the porosity. [48] Transmission line models have been used to assess the geometrical tortuosities of LIB electrodes from the fitting of their parameters to Electrochemical Impedance Spectra. [49] However, electrodes are tri-dimensional objects and therefore they have three geometrical tortuosity values, along the cartesian directions X, Y and Z. Very different values of these three tortuosities would mean a highly anisotropic electrode. Such anisotropy can lead to heterogeneities of the battery operation. [45] Evaluating such tortuosity anisotropy is of major importance.
Commercial software such as GeoDict [50] and freeware like TauFactor [51] allow evaluating such tortuosity values by solving steady state Fick's law numerically and by comparing the solution to the analytical one. However, these numbers are averaged. We believed that designing a tool that can transform a student or a researcher in a "ion" moving in the electrolyte filling the pores can be really interesting to apprehend the anisotropic character of the tortuosity and better analyze its impact on the overall electrode operation principles.
The Tortousity VR serious game proposes to the user to fly along the thickness of electrode mesostructures in order to calculate their geometrical tortuosities ( Figure 8). The user can choose among X, Y and Z directions by holding one of the HTC Vive controllers. At the beginning the user is facing the side "X" of the electrode. When the controller is placed on "Y" or "Z" circles, the electrode turns to face its "Y" or "Z" sides to the user. When ready, in order to start to fly, the user places the controller on "start". Once flying, when the user extends his/her arms he/she accelerates, while when bringing back the arms to his/her chest the user slows down. If the user touches the solid material constituting the electrode she/he will be back to the starting point with the possibility of choosing the same direction or another one to fly along. In order to find his/her way (in case of finding an impasse), the user has also the right to fly back (towards the starting point) but no longer than 10 seconds.

Nanoviewer VR serious game
Nanoviewer VR serious game was developed by us in 2017 to allow a user to manipulate in an immersive environment a wide diversity of digitalized battery-related objects such as electrolyte molecules, active material or solid electrolyte crystals and composite electrode mesostructures (Fig. 11). The goal here is to ease their visualization and assessment of asymmetries, anisotropies, and spatial organization of the materials in three dimensions. Nanoviewer VR is the precursor of Crystal VR and Tortuosity VR, but it allows manipulating any kind of digital object without other purpose than the visualization. Molecules and crystals can be originated from any kind of software for atomistic and molecular analysis and calculations, such as LAMMPS [52] or VESTA, [53] provided that the output files from these software are recorded in an atom-type file format. The composite electrode mesostructures can be originated from tomography characterizations or by using a physical model simulating their manufacturing process. The latter model, already reported by us, [54][55][56][57] is supported on a Coarse Grained Molecular Dynamics (CGMD) approach simulating LIB electrode slurries and the resulting electrode mesostructures upon the slurry solvent evaporation. By using CGMD, we have calculated several electrode mesostructures resulting from several formulations quantified by the weight ratio between LiNi1/3Mn1/3Co1/3O2 active material particles and carbon-binder, and the LiNi1/3Mn1/3Co1/3O2 particles size distribution. [54][55][56] Figure 11 shows • enlarge the digital object, using a movement of arm spacing, with both handles, to be able to "dive" inside the materials, composite electrodes, etc.; • shrink the material or composite electrode mesostructure, using a movement of arm spacing, always with both handles; • using a single handle, grab the material or electrode mesostructure to rotate it, move it in space, move it away or move it closer together.  (Figure 13). The spatial organization of these materials in a composite electrode determine its practical properties: for instance, the surface area of contact between pores and active material determine the power density of the composite electrode. Indeed, as parts of active material covered by CBD will remain electrochemically less active towards Li insertion or de-insertion, despite CBD may contain some microporosity, the path for Li + to move through can be significantly tortuous. [56,58] Other practical properties of LIBs will be affected by such interfaces, such as the cell durability and the safety. [59]   A student using the Nanoviewer VR serious game. What is seen in the VR environment is projected in a screen to allow others to see as well.

The Great Li-Air Escapade VR serious game.
As mentioned in the Introduction, LOBs, also called Lithium Air Batteries, have attracted a significant interest for electric transportation, in view of their high energy density and high theoretical capacity, even though the aforementioned electrolyte stability and rechargeability issues make them very far away from commercialization. [60] A typical LOB cell consists of an anode   Li + , O2(sol) and LiO2(sol) in the formation mechanism of Li2O2. This includes also the pore wall passivation by Li2O2, pore volume clogging and electronic tunneling effects as described in Ref. [63].
The latter captures the fact that the passivating film of Li2O2 can still conduct electrons by quantum tunneling until reaches a thickness of 10 nm for which it becomes non-conductive. The kMC model also captures the heterogeneous Li2O2 formation along the pore channel length: Li2O2 forms closer to the O2 inlet, because typically O2 transport is slower in LOB electrolytes than Li + ions. [65][66] Examples of calculated Li2O2 distributions at two depths of discharge and for different pore radius are reported in Figure 17. It can be noticed that the pore with 10 nm radius becomes totally clogged at the end of discharge whereas larger pores not, because Li2O2 film is assumed to stop growing when it reaches 10 nm thickness. For the larger pores it is underlined that the film passivation leads to the full discharge of the LOB.
The lookup table procedure is adopted because it is hard to implement real-time kMC simulations in the game without penalizing the overall computational cost of the VR environment.
The lookup table has been built by running offline kMC simulations, which allows extracting the pore geometry evolution as function of discharge history and pore size. With employing this lookup

The Great EV Escapade VR serious game
The  previously reported by us. [56,64,67] In the LSB case, the (carbonbased) cathode evolution representation is more complex ( Figure   21). In a LSB cathode upon cell discharge, solid S8 initially present in the cathode, first dissolves in the electrolyte, and then it is sequentially reduced to polysulfides S8 2-, S6 2-, S4 2-, S2 2and S 2-.
Li2S2 and Li2S form and precipitate in the cathode pores dealing to their clogging and the pores walls passivation slowing down the S8 reduction kinetics. In the game, sulfur particles initially present in the cathode pores are represented as cubes in yellow color randomly localized, and the precipitates are represented as green spheres forming gradually and randomly locating on the carbon surface upon the LSB cell discharge. Pre-calculated lookup tables using kMC simulations can be used to describe sulfur particles dissolution, polysulfides formation and Li2S precipitation. [67] The Li + intercalation process in a LiNi1/3Mn1/3Co1/3O2 -based cathode upon the LIB discharge is represented by a gradually and spatially localized changing color in the cathode mesostructure (ranging from green -pristine cathode-towards red -fully intercalated electrode-). In the LOB case, the Li2O2 distribution in a carbon matrix-based cathode is arising from a random location algorithm. The

Smart Grid MR serious game
A smart grid is an electricity grid enhanced by information technology for interlinked and automated electricity generation, transmission, distribution and control. [68][69] It encompasses electricity sources (e.g. renewables), energy sinks (e.g. EV recharge stations, houses, industry) and stationary energy storage to ensure a continuous electricity provision when intermittent renewable energies are implicated ( Figure 24). A smart grid operation can be seen, within the scope of the game theory, as a collective game involving cooperation (e.g. ensemble of electricity generators) and competition (e.g. electricity provision vs. consumption), from which complex behavior emerges. [70][71]  In order to ease the understanding by students of the interplay between electricity generation, distribution, storage and consumption, in 2019 we have designed and developed the Smart Grid MR serious game. This serious game proposes a collaborative platform mixing VR environment with real objects, constituting a "mixed reality" concept (for which the acronym "MR" stands for) ( Figure 25). also has to collect gifts, which randomly appear in the virtual map.
In contrast to the EV serious games described above, in here the number of presents is unlimited, therefore the player goal is collecting as much as she/he can before she/he loses. More gifts are collected, more difficult becomes the game as the number of trucks and the frequency of appearance of barriers become higher.
The player can lose if she/he is captured by the trucks or if the battery in the EV becomes fully discharged. To avoid the latter, There is also a virtual tidal power station in the VR environment, visible on the 3D map, that creates continuous energy from the nearby sea shore. This station is therefore not activated by the players, but it can be turned off through a Python script.
The solar panel cannot produce electricity when it is raining in the virtual city and vice-versa. The degree of production of electricity through the three real devices can be followed through three gauges (top right in the Figure 26b). If an energy source reaches the orange zone of the gauge, the energy produced is not injected anymore in the stationary battery during 10 seconds.
The worst scenario for the EV driver would be that her/his battery and the stationary battery are fully discharged: in that situation, the player will also lose the game as she/he cannot recharge the

Usage and impact
Since their creation, we have been using our serious games as Therefore, the added value of these VR games is twofold: to experiment in VR with the concepts seen in the lectures in order to validate the acquired knowledge and to identify through the serious games the students having difficulties to make them to revise the concepts. The immersive and interactive VR characters were also shown to allow students to better understand threedimensional concepts (such as crystal symmetries or electrodes anisotropies). [40] We have also studied in deep how Nanoviewer VR impacts learning performances and motivation by students. [41] For that purpose, psychology undergraduate students (i.e. no batteryspecialists), were split in two groups, one using VR (experimental group) to visualize and interact with a LIB electrode mesostructure (cf. Figure 13), and one having access to images under different perspectives of the same electrode printed on paper (control group

Conclusions
In this Concept we presented six immersive and interactive VR Even if we did not find any player who complained about cybersickness or mental overloading while playing with our VR serious games, in the next future we are going to study these aspects in more detail in order to optimize even more our serious games ergonomics. We also think that VR can help to overcome inequalities in the education system. Such inequalities affect students with social communication issues, interaction difficulties and/or cognitive and socio-emotional characteristics that require special pedagogical practices: it will be interesting to study the effect of these VR games on those students by considering differences in perception due to gender, autism or high potential.
Other perspectives of extension of our VR-related work, includes the combination of VR with Artificial Intelligence for personalized training. The VR concept introduced in this article also paves the way towards digital twins of batteries and processes (e.g. battery manufacturing plants), as wells as to a next generation of computational tools making computational modeling research accessible to a wide spectrum of researchers (including the non-expert ones), beyond also batteries. The idea of performing battery research by "playing" comforts the famous Albert Einstein's quote which says that "play is the highest form of research". [78]