Now let's move on to a direct comparison between Move AI Gen 2 and Xsens, one of the leaders in the traditional mocap market. Let's take a look at the strengths and weaknesses of each solution, as well as areas where Xsens is still out of the competition and how it can maintain its position in the AI era.
Move AI Gen 2 vs. Xsens: comparing solutions
Benefits of Move AI Gen 2
The absence of a costume and sensors is one of the key differences between Move AI and traditional solutions like Xsens. The actor doesn't need to wear anything other than normal clothing, which immediately removes a lot of restrictions. There are no pressure straps, no sensors, and no risk of anything getting knocked off in the process. This increases comfort, especially during long sessions or complex choreography, and gives the actor more freedom of movement.
For production, this is also a plus: there is no need to spend time changing clothes, calibrating and checking equipment before each session. In Xsens, all of this takes an average of 10-15 minutes per actor. With Move AI, you can just turn on the cameras and start shooting.
In the Sony Music case study, this played a key role - it was the lack of a mocap suit that allowed the team to meet the tight deadline. When the scene is set and the actor is in the shot - it's important not to waste precious time getting the tech ready. Markerless grip gives you the opportunity to work “live” and react faster to edits.
The lower hardware threshold is another major advantage of Move AI. For stable operation, of course, you need cameras with decent quality, but this is massively affordable. Many people already have suitable smartphones, and buying 2-4 simple cameras is many times cheaper than a full Xsens system.
The cost comparison here is clear: a full-fledged Xsens with gloves, multi-user support and service costs tens of thousands of dollars. Whereas an iPhone and a subscription is essentially all it takes to run Move AI. This greatly lowers the barrier to entry, especially for indie developers, small studios, educational projects, and freelancers who don't have the budget for a professional mocap.
The ability to build a working system for a few hundred or thousand dollars instead of tens of thousands - makes the markerless approach much more affordable.
Scalability to multiple characters is a strength of the markerless approach and one of the practical advantages of Move AI Gen 2. Unlike inertial systems, which require a separate suit for each participant, markerless capture is originally designed for multi-tracking. Gen 2 has taken this to a working level: under good conditions, the system can track up to 20 people simultaneously.
For Xsens, this means: two people - two sets, ten people - ten sets, with the associated costs, maintenance and synchronization. This immediately makes large-scale scenes - with extras, dance numbers, group actions - almost unrealizable without a serious budget and crew.
With Move AI, it's easier: just add cameras to cover the scene, and you can shoot a group in one take. This is what opens up opportunities for theaters, choreographers, small productions, and educational studios that want to work with multiple actors but can't afford a fleet of equipment.
Seeing and recording objects in the environment is an important difference between Move AI and inertial systems like Xsens. In Xsens, only body movements are recorded because the system tracks the position of sensors attached to the actor's body. Anything outside of this pattern - props, costume, environment - requires separate animation or additional trackers.
Move AI, in contrast, analyzes the entire video stream. This means that the capture process takes into account not only the actor's movements, but also interactions with objects - be it a sword, a chair, the edge of the stage, or even parts of the costume. While the system doesn't produce separate animation tracks for objects, the visual information about them itself is preserved. This gives context - making it easier to storyboard, animate interactions and understand the dynamics of the scene.
If an actor holds a prop or touches an object, in Xsens it has to be configured separately, either manually or via additional sensors. In Move AI you don't need to adjust anything - everything in the frame becomes part of the analysis, and this greatly simplifies post-processing and working with references.
Catch nuances of body facial expressions - another advantage of the markerless approach, which is often mentioned by Move AI users. Due to the fact that the system analyzes the entire video stream, it is able to catch micro-movements, vibrations, shifts in the center of gravity and other subtle changes in pose. For example, a slight shake of the head, a slight tension in the body, or an implicit movement of the shoulders can all make it into the final animation because the camera sees it.
In contrast, Xsens focuses on fixed sensors, and any fluctuations below a certain threshold can be interpreted as noise and smoothed out. This is especially noticeable when the system is running aggressive filtering or without gloves - small finger gestures or fleeting movements are simply lost.
As a result, animation from Move AI often looks more lively right out of the box, without the need for additional detailing. The algorithm not only reconstructs the pose, but also retains the plasticity that makes movements feel natural. Yes, Xsens gives high accuracy and stability, but it only captures what is transmitted through the sensors - no more, no less.
The lack of magnetic field and drift issues is a serious advantage Move AI has over inertial systems, including Xsens. Inertial mocap uses magnetometers for orientation, and any metal objects, electromagnetic fields, or irregular surfaces can introduce distortions. Yes, Xsens partially compensates for this with software, but if you shoot, for example, next to a car, in a hangar or in a set with metal, there is a high risk of the character drifting or “spinning”.
Markerless-system is not afraid of such interference - it analyzes the visual position of the body in each frame relative to the cameras, rather than relying on sensor readings. Therefore, in complex locations in terms of magnetic noise (ship, factory, studio with a metal floor) Move AI can show even more accurate results than the suit, which begins to “go crazy” from background noise.
Plus, inertial systems have a tendency to accumulate errors when moving for long periods of time - especially if the character walks for a long time or rotates without pausing, the algorithm can “lose” the reference point. This is usually solved by fixation inserts (e.g. put the actor on a chair to recalculate the position). The markerless approach has no such problems: it is rigidly tied to the world coordinate system of cameras, and if the character stepped forward and then returned, he will be exactly in the same place in the animation.
Remote and hidden capture is another area where the markerless system opens up new possibilities. With cameras, you can capture from afar, live, without attracting attention. Theoretically, you can record a reference of a person's movement in a natural setting (with their permission) and then translate this into animation. Putting on an Xsens suit and going out into the urban environment will not work - it's too bulky and noticeable. And Move AI allows you to capture movements even in semi-documentary format - it can be useful for reference shooting, storyboarding, choreography, research in behavioral plasticity.
The quick repetition of takes also becomes an advantage. If the scene doesn't work out, the actor just does the take again. There is no pause for reconnecting sensors, checking the costume, eliminating drift. Cameras work continuously, the system fixes, and the shooting does not stop. This is especially important on dense shoots where you need to shoot dozens of scene variations - markerless is perfect for this.
Cost for small volumes is one of the most practical arguments. If the project is a short clip, an animation of a few minutes, an experimental scene - Move AI is much cheaper. You can simply use a subscription or even a free trial. Xsens, on the other hand, has no cheap entry: either buying an expensive kit or renting a studio for impressive sums. For freelancers, indies, and educational projects, this is a key factor.
It is clear that not all advantages are universal - much depends on the context. But in the aggregate, it's clear: Move AI offers a flexibility and accessibility that wasn't available before. Minimal training, easy customization, and the ability to scale for masses or quick takes - all of this is changing the approach to working with motion. As a result, many people are reconsidering where traditional mocap is really needed and where markerless technology can do the job - without sacrificing quality, but with big savings in time and resources.
Drawbacks and limitations of Move AI Gen 2
Despite the pros, Move AI Gen 2 is not without its compromises, especially when compared to systems like Xsens, which have undergone years of fine-tuning and have been in production for a long time.
Dependence on the shooting environment is a major factor. Xsens works independently of the environment: all you need to do is put on a suit and you can shoot in a dark room, a sunny park, or a cramped closet. Move AI requires a prepared location. You need light, free space, and properly positioned cameras. If the room is cramped and you can't position the cameras from the right angles, the quality drops. If the lighting is low or with strong backlighting, the algorithm can “see” the body poorly. So markerless shooting requires more work with the environment - not just “get up and go” as in the case of Xsens. You have to think about angles, remove unnecessary objects, think about exposure - otherwise the data will not work.
Occlusions and complex interactions are the second major weakness. As soon as body parts start to overlap - for example, an actor covers his face with his hands, lies down on the floor, hugs another actor - the markerless system starts to lose its orientation. If two people are fighting or just in close contact on stage, and one overlaps the other, Move AI can start to get confused: which joint is whose, where is the arm, where is the torso. In Xsens there are no such problems in principle - the sensors “ride” with the body, and even if the hand is behind the back, the data continues to be written. The same applies to falls, rolls, grabs: in markerless shooting such scenes are a risky scenario, while in Xsens they are quite routine.
For example, if two characters fight, it's almost guaranteed to cause glitches for Move AI. And with two Xsens suits, you'll get the full trajectories of both actors with no gaps - even if they overlap. Yes, the suit won't give information about the point of physical contact (bump, touch), but the movements of the bodies will be captured accurately. Markerless-system for such a reliable capture still lacks stability in case of loss of visual contact.
Sensitivity to sudden movements is another vulnerability of markerless technology. Inertial systems like Xsens operate at a frequency of up to 240 Hz and can easily detect sudden accelerations and short-term impulses, e.g. a boxing punch, a jerk, a somersault. In most cases, video is recorded at 60 fps, and if the movement is too fast, the limb may “blur” in the frame, and the algorithm will get confused: the arm has already moved forward, but the pose in the frame - lags. This leads to inaccuracies or artifacts, especially in fast choreography or fast action. In such cases, Move AI animations often have to be manually corrected, while Xsens data is initially “clean”.
The lack of instant feedback in standard mode is also critical. With Xsens, animation is streamed to the screen in real time: the director can immediately see if an actor, for example, has passed through an object, and can immediately give a command to repeat it. This is convenient and saves a lot of time, especially on the movie set. Move AI normally works offline - you need to wait for processing to see how the scene went. And if the tracking failed or the actor went out of the camera zone - you will find out about it after the fact. The ability to stream animation is available in Move Live, but it's a separate system and is still under development. In practice, this means that AI capture requires more insurance: duplicates, tests, backup plans.
Difficulties with street and non-standard locations is another area where Xsens wins. The suit can be used anywhere: in the woods, in a car, on an outdoor stage. No cameras, no equipment set-up - put it on, turn it on, write. With Move AI everything is different: for an outdoor shoot you need cameras, power, tripods, normal light, sometimes internet for synchronization. All of this becomes a logistical challenge. Even trivial rain or wind can derail a shoot. Markerless works well in a controlled environment, but its mobility is limited. Xsens is more reliable in this respect: if the actor can physically move there, the scene will be recorded.
Limitations on the duration of continuous filming is another practical point. Theoretically, Move AI can record as long as you want, but in reality everything is based on the size of video files, ease of uploading and processing. Very long takes are gigabytes of data that take a long time to process and are difficult to debug. In addition, if multiple cameras are shooting, over time, timing drift may appear - cameras lose synchronization, and this affects tracking accuracy. With Xsens, there are no such problems: the system writes the data stream, automatically splitting it into parts, and the entire session remains consistent, without loss of synchronization. If you want 30 minutes of uninterrupted movement, go ahead. The main thing is to keep the sensors in place, but if everything is fixed correctly, you can record even an hour.
Accuracy and detail of hands and face is also a noticeable difference. Move AI is now focused on the body - it doesn't track fingers and face. For that, you need to connect separate solutions: Live Link Face for the face, Leap Motion or other systems for the hands. Xsens doesn't have built-in facial expression capture either, but it does have integration with Manus gloves that allow you to capture every finger. On the plus side, Xsens accurately tracks hand and foot orientation, and that's especially important in complex scenes.
Move AI, on the contrary, in the case of overlapping hands or a strong tilt of the hands, errors may occur - the system simply does not see how exactly the palm is turned and starts guessing. And fingers remain a “blind spot” - they are simply not in the grip. If you use AI for hands, you need to capture separately, from a different angle, and run other models. This is possible, but adds layers to the piplane.
In the end, if the task is to capture body, hands, face in one frame, with high accuracy and without gaps - Xsens in conjunction with gloves and Faceware gives a more comprehensive solution, although it costs more. Move AI can be augmented with third-party systems, but out of the box it's still only body-centric.
Debugging and predictability is an important practical aspect. Xsens is a tool with clear, fine-tuned behavior: if something goes wrong, it's usually clear why. Dead sensor? Lost signal? A calibration failure? It's all solvable, and the solutions are documented. Move AI, on the other hand, works like a black box: if the algorithm suddenly made a mistake in a certain pose or the tracking “went wrong”, it's almost impossible to immediately understand what's wrong. Users have noted that they had to re-record takes with different lighting or camera positions, experimenting to get a stable result. This can be critical if there's tight timing on a shoot. The suit either writes or it doesn't - and if calibrated, it will produce a repeatable track. An AI system, on the other hand, can interpret the same scene slightly differently depending on nuances that aren't always obvious.
Cost at high volumes is another important factor. If a project lasts for months and involves hundreds of hours of capture, the Move AI subscription model can end up being more expensive. Xsens requires an investment at the start, but then you can write indefinitely without paying extra for meterage or number of sessions. The hardware lasts for years, and its amortization is clear. In addition, Xsens data is stored locally, and the studio does not depend on a third-party service. Move AI, on the other hand, is a cloud-based tool: there is a risk of changes in policy, pricing, or even shutting down the service. It's not a problem for now, but for long-term planning, such dependence can be sensitive. Some are waiting for an on-premise version of Move AI - local processing with no attachment to a platform - but for now this is in the future.
To summarize, Move AI Gen 2 is inferior to Xsens in terms of reliability, predictability, and versatility in challenging conditions - in combat, on offsite locations, and in unstable shooting environments. But for most standard tasks, especially in digital production and interactive media, its advantages - flexibility, no costumes, easy start-up, scalability - make its disadvantages not critical. It all depends on the context. If you need a guarantee of results under any conditions, Xsens remains the only alternative. If accessibility, mobility and fast iteration are important - Move AI becomes a powerful alternative, which in many scenarios can already replace classic mocap. Below is a breakdown of tasks where Xsens is objectively better.
Xsens advantages (and areas where it leads)
Xsens (now part of Movella) is one of the pioneers of inertial mocap and a recognized leader in the category. Their flagship solutions - Link and Awinda - are used in movies, games, sports science and research. Despite the emergence of markerless alternatives like Move AI, Xsens continues to hold its ground, largely due to a number of unique technical and practical advantages.
Independence from outdoor conditions is a key reason why Xsens remains an indispensable tool in outdoor filming. The costume doesn't require cameras, lights or a large area - it just needs the actor to be able to move freely. That's why Xsens is used on location: you can record scenes in the woods, outdoors, in the mountains, without setting up equipment around it. A Markerless grip for such tasks requires a camera, power, a stable platform and a precisely built configuration - you can't just go out into the field with it. In sports projects it is also critical: Xsens allows you to shoot running, jumping, movements on non-standard trajectories (for example, 100-meter race), where video cameras physically can not keep up with the actor or can not be installed at the right distance.
Resistance to occlusions and complex interactions is another strength of Xsens. If two people in costumes are hugging, fighting or tumbling - each costume writes data independently and the system doesn't get confused. There's no risk of one person's track merging with the other's, as sometimes happens in markerless-capture when bodies cross. Yes, Xsens will not determine the point of contact between the actors - that's left to the staging level - but the movement of each body will be captured accurately and consistently.
Falls, jumps, and rolls are no problem either. Sensors are attached directly to the body, and no change of angle, loss of visibility, or overlap affects the system. An AI system in these conditions can lose body parts due to overlaps or out of camera range, and Xsens continues to read angles and positions.
In the end, when it comes to reliability in complex movements, especially in stunt and fight scenes, Xsens remains a benchmark of stability that markerless simply cannot compete with yet.
High frequency and blur-free operation is an area where inertial systems are still firmly in the lead. Xsens operates at a frequency of up to 240 Hz, which gives extremely accurate registration of accelerations and rotations. The system is image-independent, so there is no motion blur. The data is always clear - even if the actor moves lightning fast, even if he hits with a fist with a sharp acceleration, the suit captures it all in real time and with the necessary detail.
AI systems, including Move AI, work with 60 fps video - and if movement occurs between frames, these nuances are lost. Example: a fencer makes a sharp lunge, but the video just shows a blurry hand. The algorithm has to guess the trajectory, and this gives errors. Xsens gives the exact trajectory of the hand in such a situation, because it reads the motion directly from the body. This is why inertial mocap remains more accurate and complete for fight scenes, stunts, acrobatics, and high-speed choreography.
Instant previews and interactivity is another benefit that's hard to overstate. With Xsens MVN, you can stream real-time data into Unreal or Unity and see how a character is moving right away. This changes the approach on set: the director sees the result right in the frame, can instantly understand if the movement is working, if the action matches the virtual scene, if it needs to be reshot. The actor sees his avatar. The operator can customize the virtual camera to match the live performance.
In virtual production or live shows, this is a must-have: the character is animated right during the performance, and the audience sees the digital hero in real time. Move Live is just starting to reach this level, while Xsens has been perfecting this functionality for years - stable streaming with minimal latency, compatibility with major engines, and proven pipelines.
Even the movie industry is increasingly using Xsens for virtual LED scenes: an actor in costume, his image immediately appears in 3D space, and the entire crew can work in conditions as close as possible to the final shot. So far, markerless capture only aspires to this - there are no alternatives to Xsens in real-time applications.
Data consistency and repeatability is one of the criteria for which Xsens is particularly strong. After calibration, the system is rigidly tied to the set parameters: if the character's height is set as 1.8 meters, it will remain that way in all takes. Each movement is recorded in physically meaningful coordinates, and each unit of data (angle, acceleration, orientation) is the result of direct measurements from sensors.
In Move AI Gen 2, the scale is also calculated, especially in multi-camera calibration, but there may be slight fluctuations between takes - the character is slightly higher, or slightly lower, if the scale of the scene was not perfectly defined. This is not critical for creative tasks, but in analytics, biomechanics, medicine or sports, where absolute values of angles, step lengths and accelerations are important, Xsens is more reliable and accurate. It is no coincidence that it is used in universities, laboratories and research institutes all over the world - where you need guaranteed results validated in numbers. The Markerless approach is just starting to penetrate there and still needs to be validated.
Comprehensive offering and support is another area where Xsens wins. It's a mature product with a whole ecosystem built around it: tech support, community, regular updates, documentation, custom configurations. Studios know that they can not only buy a suit, but also get consultation, technical support, help with integration, and on-site visits. Move AI is still working on a more “product” basis - self-service, subscription, online documentation. This is fine for small teams, but on big shoots and important projects you often need personal support, and here Xsens gives you confidence.
Producers and production managers are more likely to choose a proven solution - especially when dealing with clients, equipment rentals or tight deadlines. Many mocap studios rent Xsens along with equipment and operator - and it's standardized. There is no such infrastructure for Move AI yet.
The integration of gloves and additional sensors is another practical advantage of Xsens. The system is easily expandable: you can connect Manus gloves to capture finger movements, add sensors on props (e.g. swords or weapons), and it will all work in a single environment, synchronously and accurately. The data is collected in MVN and exported as a single mocap stream. A Markerless system can't do that: it can't track fingers or an object in your hand directly - that requires a separate video, a different neural network, complex customization, and in the end it still won't be part of a complete system. So if you need full body + finger + object capture, Xsens is the central hub that brings it all together.
Reliability over long periods of time is another area where Xsens remains in the lead. If a project is stretched out over months and capture is happening every day, the system needs to be predictable and resilient. Xsens suits are exactly that - the sensors are designed for wear and tear, the software is stable, updates are regular, and support is lined up. Move AI in this respect is dependent on third-party factors: cloud processing, server utilization, internet stability. If something goes wrong, a session can be delayed or even disrupted. In addition, large studios or companies with strict security requirements may prohibit the use of cloud services. Move AI simply won't work there, while Xsens is fully autonomous, everything works locally.
In the end, Xsens remains the best choice when stability, accuracy and predictability are the priority. It wins in challenging environments, long and regular surveys, real-time, projects where you need full control over the data or where the cost of error is high. That's why Xsens is used in Star Wars-level projects to capture battle scenes with dozens of actors, in professional sports to measure and analyze techniques, and in scientific research where every degree of angle must be accurate. In such cases, AI solutions still fall short in terms of stability.
Where Xsens remains the best tool
Let's highlight some of the specific areas where Xsens currently holds the lead over Move AI Gen 2:
Movie stunts and fight scenes. On action sequences, especially those involving stuntmen, Xsens is almost always the choice. The reason is simple - it works reliably even in chaotic movements, collisions, falls, rolls, and explosions. Markerless-system in such conditions can lose tracking: if the body is partially covered, if the actors clashed in a fight, if someone fell on the ground - Move AI will not always cope. Xsens has no such problems: each sensor stays with the body, and the scene is recorded in its entirety, without skips.
In addition, in stunt scenes, the ability to instantly visualize is important. The director wants to see the CG character right on the set to adjust the staging. This is only possible with a suit that gives real-time data. Markerless in such tasks is so far a compromise with latency. So for action scenes, where it often takes one take to shoot, the choice is obvious: Xsens delivers results immediately and without surprises.
Live performances and broadcasts. When an artist takes the stage in costume and their avatar is broadcast in real time - for example, as part of a concert, presentation, or stream - it's critical that there are no delays, skips, or instability. Xsens has become the industry standard here: it's used for virtual presenters, avatars, Fortnite shows, and other large-scale projects. It's stable, gives accurate live tracking, and integrates well with Unreal, Unity, Omniverse, and other platforms.
Move AI is just starting to take this direction. Move Live has potential, but it doesn't show the same stability yet and is not used in large live projects. So if you need to run a show right now, without risk - take Xsens. It is proven, the technicians know it, and it just works.
Sports analytics and medicine are areas where Xsens has no competitors yet. It's not just animation that matters here, but metrically accurate data: angles, accelerations, segment lengths, and symmetry of movement. Xsens provides verified measurements based on real sensors calibrated for a specific person. Markerless systems, including Move AI, do not yet have a sufficient degree of confidence - too much depends on shooting conditions, video quality, and model interpretation. When diagnosing gait or analyzing rehabilitation, patients cannot risk data errors. In such tasks, AI capture is still an experiment, and Xsens is a practice tool.
Plus, Xsens can be used outside the studio, in the sun, in a natural environment, which is important for analyzing running, for example. Markerless shooting in such conditions is faced with over-lighting, contrasts, shadows, which worsens the result. For serious biomechanics, rehabilitation, prosthetics, and sports research, Xsens remains the only alternative solution, and Movella is actively developing this area.
Long takes and shooting outside the studio is another advantage. Move AI, being a video-dependent service, is limited by file sizes, camera stability and synchronization. Recording a continuous hour of dancing or practicing is risky. One glitch and the entire session could be ruined. Xsens, on the other hand, writes a stream of data in real time, with no risk of loss. Its software can handle long sessions, and this is especially valuable in choreography, theater, and experimental productions. In addition, neural network datasets for motion training are often created precisely using Xsens - because of the stability, cleanliness and duration of the recordings.
The individual characteristics of the actor are also important. Xsens is calibrated for a specific body: height, limb length, proportions. All data precisely matches the physique of a particular person. Move AI calculates the skeleton from video, which gives an approximate result based on an averaged model. For tasks where you need to preserve the unique movement mechanics of a particular actor or athlete, a suit is preferable.
Privacy and control are a separate factor. Xsens works completely locally: no data leaves the studio. This is important in closed projects, movie sets, and companies with security restrictions. Move AI requires uploading video to the cloud by default. Even if the security is at a good level, the very fact of dependence on an external service may be unacceptable for some clients. Until a full on-premise version is available, Xsens maintains a strong advantage here.
Thus, Xsens remains a key tool where stability, control and precision are important. It is the tool of choice in complex stunt filming, for live performances, in sports, medicine, biomechanics and R&D. It is not replacing AI grip - nor is it being supplanted by it - but is taking its rightful place in the professional Pipeline, where the cost of error is too high to experiment with.
Perspectives: How Xsens can stay on top amid AI solutions
The development of AI-mocap poses a serious challenge to traditional manufacturers: how to stay relevant and not lose ground. In the case of Xsens (Movella), there are several strategies that can keep them at the top amid the rapid growth of markerless systems.
Integrating AI into their own products. Xsens doesn't necessarily have to compete with AI solutions directly - on the contrary, it can use the same technologies to its advantage. One obvious way is a hybrid approach: a combination of computer vision and inertial sensors. Cameras could be used to compensate for drift, refine position in a world coordinate system, or capture a markerless face. Prototypes already exist where multiple cameras complement the suit, addressing the weaknesses of both methods. If Xsens realizes such a system - reliable, accurate, occlusion-free, and fully georeferenced - it could surpass both purely inertial and purely markerless solutions. There are also rumors of possible AI plug-ins from Xsens - for example, automatic mocap cleaning using a trained model. This is a logical step: don't fight AI, but build it into your stack, strengthening your own product.
Focus on strengths: reliability and real-time. Xsens relies on what it is already good at. Reliability, repeatability, operation in all conditions and live motion broadcasting are where inertial mocap continues to excel. Related products are evolving: Xsens DOT - miniaturized sensors for motion tracking in AR/VR, better software, easier calibration, less dependence on the operator. The easier a suit becomes to use - the less reason users have to switch to markerless. If tomorrow there is an Xsens suit that can be put on in a minute and calibrated automatically without connecting to a computer, it will be a strong argument to stay with a system that is already trusted.
Additionally, Xsens can play on price flexibility. They already have Xsens Animate, a software subscription without having to buy the full suite outright. If the company starts offering different ownership models - rental, subscription, educational licenses - it will open the door to new customers without lowering the status of the technology.
In this way, Xsens can not just survive in an era of AI takeover, but reformat its offering by strengthening strengths, eliminating weaknesses, and embedding AI as part of the solution.
New application areas. While AI-mocap is being actively promoted in the media, movie, and gaming industries, Xsens can delve into niche B2B segments where the priority is not visual expressiveness, but high accuracy, certification, and reliability. These include medicine, clinical biomechanics, sports analytics, military simulators, robotics, and ergonomics. In these fields, accurate numerical measures that can be interpreted, repeated, validated are important. Markerless-technologies are not yet trusted in such tasks - there are too many uncertainties. Movella is already betting on it: Xsens is increasingly appearing in university labs and medical projects. If the company strengthens its position in these areas, it will be less dependent on the fluctuations of the entertainment market, where today the hype is high and tomorrow the decline.
Service and ecosystem. Xsens can strengthen its position by offering a premium level of service. Instead of just selling equipment, the company can offer a complete package: equipment + training + implementation + support. In general, this is already working. Markerless systems have a largely self-service model: download it, figure it out for yourself. But in production, you don't always have the time and resources to experiment - they value reliability and support. If Xsens continues to develop a global network of partners, expands technical support (ideally 24/7), and offers flexible solutions to meet client needs, it will become a strong competitive advantage, especially for large studios and international projects.
Integration with graphics engines. Xsens already works well with Unreal and Unity, but you can go further - offer ready-made production sets: costume + cameras + scene templates + bundle with LED screen or virtual camera. Something like “Stage in the Box” - a boxed solution for running a virtual stage without additional integrations. This approach not only saves customers time, but also moves Xsens from being a “tool” to a “complete platform”. This adds value that pure AI services do not yet offer.
Focus on the quality of the result. Xsens can clearly communicate the difference in quality in the marketing strategy. For example: "AI delivers impressive results, but can fail at key moments - occlusions, fast turns, falls. Xsens is stable and clean at all times." Illustrations of such comparisons, real cases and examples of AI system errors can convincingly show that professionals need a tool that can be relied on without reservations. In addition, quality guarantees can be introduced: for example, the system can be claimed to provide more than 95% accuracy of movements in all conditions - something that no markerless service can guarantee yet. This brings the discussion back to the point: not just “cool and fast”, but accurate, stable, reproducible. And that remains a field where Xsens is objectively strong.
Overall, Xsens has every reason to maintain its leadership position if the company continues to adapt and intelligently integrate AI development into its ecosystem. The history of technology has shown time and again that new approaches do not necessarily displace old ones - more often there is a redistribution of roles or synthesis. The same thing is happening now with the motion capture market.
It is highly likely that the market will divide: markerless services will occupy the segment of mass, fast and inexpensive content - mobile applications, previsas, videos for social networks, experimental projects. But high-level production, scientific research, medical diagnostics, real time, live scenes, stunts and everything where the price of error is high will be left to suits, perhaps in a hybrid form, where AI helps, but the suit remains the basis.
Xsens, as one of the most mature players in the market, is well positioned to not only hold on to, but to channel this trend. By investing resources in AI integrations, hybrid solutions, and enhanced usability, the company can offer a product that is hard to compete with: reliable, accurate, real-time, scalable, and customizable. In this configuration, Xsens can maintain its status as the benchmark for professional mocap - and build on its technological synthesis rather than resistance to change.
Practical conclusions and recommendations
Move AI Gen 2 and Xsens - both solutions are powerful, but are customized for different tasks. Move AI Gen 2 has made mocap accessible: you no longer need a suit, trackers, studio - you can just put cameras or phones, record, upload, and get animation. This is handy for small teams, indie developers, studios without access to equipment. It's also suitable for previs and prototypes where you don't need perfect accuracy, but speed and flexibility are important. If the scene is not too complicated - one or two actors, minimum overlaps, normal lighting - the result is very good, close to what marker systems produce, but without infrastructure costs. An important point is competent shooting: you need to understand how to place cameras, how to organize the location so that tracking works properly. Gen 2 has already been implemented in production, used in gamemade, music videos, even in advertising - and it shows itself as a tool that gives freedom and shortens the cycle. It's important to keep in mind that the system is not magic - you'll have to allocate time for tests, the take may not work, you'll have to adapt. But if the approach is built up, you can achieve studio quality without spending tens of thousands on equipment.
On the other hand, Xsens remains a reliable tool, a kind of “insurance” in mocap. Especially when conditions are not ideal - scenes with falls, fights, complex interactions, real-time shooting or off-site work. In such situations, the suit gives confidence in the result: no matter what happens on the set, the system will record the movement. It is stable in case of sudden and chaotic actions, it shows at once what has worked and does not require guessing how the take will be processed. This is critical in projects where you can't afford to make a mistake - for example, in medical tasks, scientific measurements, military simulations.
If the project is large, with a serious budget, a long shoot or many actors, using Xsens reduces the risks. You get a proven solution with repeatable behavior. Also, when you need to shoot outside the studio - in the woods, in the hall, on the street - the suit works autonomously, without dependence on networks, cameras, lights. For many productions, this is simply the only reliable option, especially when stability and predictability are important.
The choice between Move AI Gen 2 and Xsens depends on the task and format of the project.
If you're working on short clips, indie games, previs or experiments, it's often more convenient to take Move AI Gen 2. It's quick to set up, doesn't require a suit, and gives you results that can be immediately loaded into Unreal or Unity. It allows you to quickly run scenes, make edits, reshoot without unnecessary steps. Everything happens quickly - minimum equipment, minimum logistics. For many tasks of this level, Gen 2 quality is already enough to use it in the final Pipeline.
If the project is serious - a movie, AAA game, stage performance or live show with many actors, complex choreography, or real-time is important - Xsens provides reliability that markerless systems do not yet provide. Where you can't lose a take because of a tracking failure, where scenes are complex, re-shooting costs are high, and the process must be predictable - the suit relieves stress. Everything is recorded consistently, the system is tested, and the team can focus on the work rather than technical debugging. Such projects usually have both the expertise and the budget to use a classic solution without compromise.
In reality, more and more often a mixed approach is used - both tools can safely coexist and complement each other. For example, in the early stages you can use Move AI for quick rough capture: set up cameras, shoot a basic scene, make animatics, coordinate storyboarding and directing. And then, when you need to finalize the episode, re-record key moments through Xsens - with maximum accuracy and full quality assurance.
And vice versa: if the costume is already used for the bulk of movements, auxiliary animations - background NPCs, variations, simple scenes - can be captured via Move AI to save time and unload the technique. Data formats are compatible, which means everything can be merged into a single piplane without losing the integrity of the project. This flexibility is a new phenomenon in the mocap market, and it works in favor of studios and production houses.
As for Xsens, its future likely lies not in head-on competition with AI, but in the evolution and integration of new technologies. You can already envision AI-enabled modules appearing in the suit - for example, auto-correction of data, a hybrid with a camera for positioning, smart filters, and neural network-based post-processing enhancements. This is not a contradiction, but a logical development. The boundary between AI and classic mocap is gradually blurring.
Today, Xsens holds the bar thanks to its reliability and deep validation, while Move AI is driving the industry forward with an innovative, user-friendly and affordable approach. In the end, users win: directors, animators, and studios have more options, iterate faster, and choose a tool to fit the task rather than budget or location constraints. This is the real progress in motion capture.
Opmerkingen