Definition: Motion capture (mocap for short) is the name for methods that measure the movements of a person or an object and record them as digital data. The result is a time sequence of positions and joint angles that you can analyze or transfer to a digital skeleton. This gives realistic motion sequences for 3D applications and reliable measured values for ergonomics.

In B2B mechanical engineering, motion capture is used for: ergonomics analyses at assembly and operator stations, training content with real hand movements, animated assembly and service instructions, design reviews with digital human models, and presentations that show the operation and maintenance of a machine.

Advantage: Real movements instead of estimated sequences, measurable and reusable in web, kiosk, AR and VR.

English: Motion Capture, Mocap, Performance Capture
German: Bewegungserfassung, Bewegungsaufzeichnung

How It Works

Basic principle

  1. Capture: Cameras or sensors worn on the body measure the movement.
  2. Reconstruction: Software calculates the positions and joint angles of a skeleton model from this data.
  3. Cleanup: Gaps, outliers and noise are corrected.
  4. Transfer: The movement is transferred to a digital figure (retargeting) or analyzed as a measurement data set.

Optical systems with markers

Several infrared cameras capture reflective or actively lit markers on the body. The 3D position of each marker is triangulated from the camera images.

  • Strength: High accuracy. Research uses it as the reference.
  • Limitation: A fixed setup with calibrated cameras. When markers are hidden (occlusion), for example when a person reaches into a machine, gaps occur.

Inertial systems (IMU)

Sensors with an accelerometer, a gyroscope and usually a magnetometer are attached to the body, often in a suit or with straps.

  • Strength: Mobile, no camera view necessary, and you can also use it directly at the real workstation in the hall.
  • Limitation: The position is extrapolated from measured values and can drift over time. Steel structures and electric motors interfere with the magnetometer.

Markerless systems

Cameras without markers detect the body posture with image processing and AI. The preparation effort decreases considerably. Studies show promising results for ergonomics analyses, but the accuracy depends strongly on the number of cameras, the viewing angle and occlusions.

Method Location Accuracy Effort
Optical with markers Lab, fixed measurement room Very high High
Inertial (IMU) Hall, construction site, customer site Good, drift possible Medium
Markerless (camera + AI) Workstation, lab Depends on the setup Low

B2B Applications

Ergonomics analysis

Assembly, setup and maintenance put strain on the back, shoulders and arms. With motion capture, you can measure body postures objectively and compare them with assessment methods such as RULA, REBA or OWAS and with standards such as ISO 11226 and EN 1005-4. This is especially valuable in the design review: the movement of a worker is transferred to a digital human model that works in an early 3D model of the machine. This shows before construction whether the operator panel, tool change or filter access is ergonomically reachable.

Design review with a digital human model

Digital human models represent different body heights. In combination with recorded movements, you can check whether a workstation works equally well for short and tall operators: reach, lines of sight to the display, bent postures when parts are removed. Alternatively, an employee with mocap equipment moves through the virtual machine in a VR scene, and the posture is recorded and assessed.

Training

An experienced technician demonstrates a work step, and mocap records it. From this, you get an avatar that shows the sequence in VR training or web training. Trainees see the correct body posture and the order of the hand movements from every angle.

Animated assembly and service instructions

Instead of step-by-step photos, an animation shows how a part is loosened, turned and inserted. Mocap gives the natural movements of the hands and the body for this. The instructions run on the tablet at the machine, on a kiosk in the workshop or as an AR overlay directly on the real part.

Workflow for animated instructions: 1. Demonstrate and record the work step on the real part or on a replica 2. Clean up the motion data and transfer it to a neutral figure 3. Synchronize the animation with the 3D model of the machine 4. Add labels, warnings and step numbers 5. Deliver for web, tablet, kiosk or AR

Advantage over video: You can view the instructions from every angle, pause them and adapt them after a design change without filming again. Language versions come from texts, not from new recordings.

Sales and trade shows

Animated operators make a machine presentation easier to understand: How is the machine loaded, where does the operator stand, how does the format change work? An immersive experience with realistic movements is more credible than a figure in a rigid pose.

Challenges

  • Post-processing: Raw data needs cleanup, especially with occlusions and drift.
  • Retargeting: Skeletons differ in proportions and joint structure. The transfer to a different figure needs adjustment.
  • Data privacy: Motion data relates to persons. For recordings at the workplace, the data privacy rules apply. In Germany, you often must also involve the works council.
  • Effort and costs: Equipment, setup and analysis need expert knowledge. For individual projects, a service provider is often more economical than your own system. For regular ergonomics studies, a purchase can be worth it.
  • Validity: A single recording shows the movement of one person. For reliable ergonomics results, you need several persons of different body heights.
  • Industrial environment: Noise, metal, changing light and tight spaces make the measurement more difficult. The location determines the choice of method.

Frequently Asked Questions (FAQ)

What is the difference between motion capture and motion tracking?

Tracking follows the position and orientation in real time, for example of a headset or controller, so that an AR or VR application can react. Motion capture records movements, usually of the whole body, to analyze them or to reuse them as an animation. The technology overlaps, but the purpose is different.

Optical or inertial: which is suitable for mechanical engineering?

For precise studies in the lab, optical systems with markers are the reference. For recordings at the real workstation or at the customer, inertial systems are often more practical, because they do not need a camera view. Many projects combine both approaches.

Do you need your own mocap studio?

Not necessarily. You can use inertial and markerless systems on the move. For individual projects, specialized studios and service providers offer recordings.

In which formats is mocap data shared?

For biomechanics and ergonomics, C3D is common. For animation, BVH and FBX are common. For web, AR and VR, the animation is usually transferred to a skeleton and delivered as glTF or USDZ.

Can you use mocap data in real time?

Yes. Many systems stream the movement live into a 3D application, for example for an avatar in a shared VR session. For ergonomics analyses and instructions, the data is usually recorded and post-processed.

Industry Standards and Technologies

  • C3D – open file format for motion and biomechanics data
  • ISO 11226 – ergonomics, evaluation of static working postures
  • EN 1005-4 – safety of machinery, evaluation of working postures and movements (DIN Standards Committee Ergonomics)
  • OpenSim – open-source software for motion analysis
  • glTF 2.0 (Khronos Group) – skeletal animations for real-time 3D
  • OpenXR (Khronos Group) – open interface for AR and VR devices

Capture and tracking: - Sensors – measurement technology in XR devices - Eye tracking – capture of the gaze direction - Markerless tracking – tracking without markers - Marker tracking – tracking with markers - LiDAR – depth measurement with a laser

Platforms: - Virtual reality – training in VR - Augmented reality – instructions on the real part - Extended reality (XR) – umbrella term - Real-time rendering – animation in real time

3D content: - 3D model – machines and human models - 3D pipeline – data preparation - USDZ – AR format with animation - Retargeting – transfer to other skeletons - Digital human model – virtual worker

Applications: - Immersion – being fully absorbed in the application - Digital twin – digital image of the system - Data privacy – handling of motion data - Ergonomics analysis – evaluation of working postures - Animated instructions – assembly and service