The Shift from Motion Metrics to Human Interaction Data
For decades, human movement has often been represented through a relatively compact set of variables: joint angles, range of motion, trajectories, repetitions, velocity, and force. These measurements remain essential in biomechanics, ergonomics, sports science, and engineering because they describe how body segments move and how demanding a task may be. But physical products do not interact with angles alone. A wearable follows a changing body surface. A seat supports tissue that redistributes under posture and load. A healthcare device must maintain contact as the user bends or rotates, while an automotive interior must accommodate a body whose occupied space changes continuously during entry, driving, and exit. For human-centered product design, movement therefore needs to be understood as a changing relationship between the body and the product, not simply as a sequence of joint coordinates.
This shift changes the type of human evidence product teams need. 3D human body data can describe surface geometry, contours, local curvature, cross-sections, asymmetry, and body proportions that linear measurements cannot fully capture. Add posture and time, and designers can begin examining how those characteristics change while a person reaches, bends, sits, walks, grips, or wears a product. The design value lies less in creating a more visually realistic animation than in understanding changing clearance, contact, occupied space, and accommodation throughout use. This broader treatment of movement also underlies Designing Workspaces Around Real Human Movement: Why 4D Motion Data Matters, where human movement becomes a design input rather than a series of isolated poses.
The Design Limits of Joint Angles and Static Anthropometry
Joint-angle data tells an engineer how segments are positioned relative to one another, but it does not fully describe the geometry of the body surrounding those joints. When a knee flexes, a shoulder rotates, or a person moves from standing to sitting, local body shape, surface orientation, cross-sectional geometry, and soft-tissue distribution can change along with skeletal posture. Those changes become consequential when a product is worn, supported, enclosed, or positioned against the body. A device may remain within an acceptable joint range while still shifting away from its intended sensing location. A brace can allow the required movement yet create concentrated contact in a particular region. A seat or support surface can satisfy dimensional requirements while interacting differently with users whose body shapes and tissue distributions vary. Geometric compatibility is therefore not the same as product-human fit.
Static anthropometric measurements remain indispensable, particularly for defining overall dimensions, clearances, reach zones, and target populations. The problem appears when they are treated as a complete representation of the user. Two people with similar stature can differ in segment proportions, body depth, local curvature, mobility, posture strategy, and the way they accomplish the same task. Products that physically interact with the body amplify these differences. Wearables must remain stable through motion; healthcare devices need reliable placement and manageable pressure; furniture must accommodate changing seated postures; vehicle interiors must support reach, visibility, ingress, and egress; and robots operating near people need models of the space humans actually occupy while moving. The design challenge is therefore to combine anthropometric data with body shape, posture, movement, and task context rather than forcing human diversity into a single average representation.
Dynamic Digital Humans as a Validation Layer
The convergence of 3D scanning, motion capture, anatomical landmarking, skeletal models, sensor data, and computational simulation is making a different kind of digital human possible. Instead of representing a user as a fixed mesh in a neutral pose, a digital human model can connect body geometry to joints, landmarks, postures, and movement sequences. Time-dependent datasets can then show how the spatial relationship between a body and a product develops across an activity. Engineers can investigate whether a hand remains within a comfortable reach envelope, whether a head or knee approaches a surrounding structure, whether a wearable changes position relative to the body, or whether a product interferes with a critical movement. As explained in Why Digital Humans Need Movement Data, Not Just Body Measurements, the useful digital human is increasingly becoming a model of human action rather than a static representation of appearance.
That distinction matters because virtual usability simulation is most valuable when it supports decisions before a physical design becomes expensive to change. Representative human models can be paired with product geometry to screen alternative dimensions, placements, clearances, and interaction conditions across different users. A wearable developer can compare fit across local body shapes and postures; a mobility team can investigate clearance throughout a complete entry sequence; a furniture designer can examine reach and posture across body-size groups; and a healthcare team can identify configurations that warrant closer physical testing. Virtual analysis does not eliminate prototypes, biomechanical measurement, or real-user trials. It changes when those resources are used. How Digital Human Simulation Validates Products Before Physical Prototyping provides the complementary principle: simulation can identify likely constraints earlier so that physical testing is concentrated on the most meaningful designs and use conditions.
Human Data Infrastructure for AI Product Development
Making this approach useful at scale requires more than collecting large numbers of scans or motion files. Human data must be structured so that body geometry, measurements, anatomical landmarks, joints, postures, movements, population characteristics, and product conditions can be interpreted together. Coordinate systems and landmark definitions need consistency. Representative models need to reflect meaningful variation rather than simply reproduce an arithmetic average. Capture conditions and processing methods need enough consistency for comparisons to remain reproducible. These requirements become even more important as AI enters product development, because fast generation creates many more concepts to evaluate. The question is no longer only how quickly AI can create a plausible geometry, but whether that geometry can be tested against credible representations of the people expected to use it.
Comfo Labs is developing this connection between human data and product-development decisions. SIZE LAB provides access to 3D human body data and analysis resources that can support anthropometric analysis, body-shape interpretation, representative 3D personas, and ergonomic design decisions. Doodll extends the workflow toward AI-assisted product development, connecting product ideation and 3D development with earlier evaluation of usability, fit, and interaction. The objective is not to replace engineering expertise or physical testing with a digital model. It is to move human evidence upstream, while product geometry can still be changed quickly and economically. As product development becomes increasingly AI-assisted, the more consequential shift may be from designing around assumed users to data-driven product development built around bodies in motion.
Korean Version:
운동 데이터의 패러다임 전환: 관절 각도가 아닌 ‘체적 변화’로 분석하는 근력
