Motion capture files
Baseline and subject motion files enter a selected product workflow rather than a generic black box.
Technology
Deterministic movement evidence engine.
Kineologist is built around deterministic analysis rather than hidden judgement. Motion capture files move through a product-specific pipeline; the backend creates phase markers, crop windows, baselines, telemetry channels, plots, reports and debug artifacts; the interface then lets the reviewer inspect those outputs.
Live pipeline spine
From motion files to reviewable evidence.
The technology spine is intentionally explicit. Kineologist separates computation from presentation: the backend generates deterministic artifacts, while the GUI and report layer render those artifacts for review. This keeps the system inspectable, repeatable and easier to test across products.
Context detector
PITCH, GAIT or a custom deployment defines the phase/event logic that makes the motion meaningful.
Phase/event crop
The movement is cropped around relevant windows before the rest of the analysis is calculated.
Baseline generation
Reference clips can be normalised and averaged into a repeatable comparison authority.
Telemetry extraction
Curves, rotations, velocities, contact channels, events and deltas are exported as evidence.
Muscle proxy sequencing
Motion-derived proxy groups are created from joint rotations, motion and velocities.
Rule engine
Research-derived, context-specific rules convert persistent patterns into deterministic findings.
Reports / plots / CSV
Reports, plots, CSVs, JSON markers and debug snapshots are written as inspectable artifacts.
Review surfaces
The interface presents the generated evidence through 3D review, TrackView, reports and exports.
Sport Pack architecture
Sport Packs let one engine operate in different movement contexts.
The Kineologist engine stays shared, while movement-specific logic lives in Sport Packs. A pack can define detector behaviour, phase/event semantics, telemetry channels, rules, thresholds, report language and review outputs for a particular context. This keeps PITCH, GAIT and future custom deployments on the same artifact-driven backend while preventing one movement model from contaminating another.
Shared engine
Parsing, artifact generation, baseline handling and review surfaces stay common.
Pack-owned detectors
Movement-specific event and phase detectors live inside the relevant pack.
Telemetry schema
Each pack declares the channels that matter for its movement context.
Rules and thresholds
Pack rules process telemetry through tunable, repeatable analysis logic.
Report lexicon
Each pack owns the wording used to explain findings safely.
Deployment scaling
New contexts can be added by creating, testing and validating a new pack.
Context-specific detectors
Different movement contexts use different detector authorities.
PITCH uses CALF-derived STRIDE authority before upper-body throw initiation. GAIT uses foot-roll context, calf-refined toe-off and tibialis-refined initial contact. Each detector path produces backend-owned event truth for crops, telemetry, phase locking and reports.
PITCH phases
Baseball pitching review is anchored by P1–P4 phase markers.
GAIT events
Walking review is anchored by IC/TO events and fixed-step logic.
Custom detectors
A deployment can define its own meaningful windows.
Comparable crops
Subject and baseline use the same crop authority.
Context-aware phase-warping
Different movements align around different anchors.
Sport packs supply their own meaningful anchors. PITCH aligns P1, STRIDE, P2, P3 and P4; GAIT aligns repeated IC/TO events. Kineologist moves the reference curve horizontally to equivalent events without rescaling its original physical magnitude.
PITCH anchors
P1, STRIDE, P2, P3 and P4 can define the pitching warp map.
GAIT anchors
IC/TO walking events can define the gait warp map.
Original timing
Native timing differences remain visible in the Original view.
Phase-locked review
Equivalent movement events align for comparison without changing the source artifacts.
Artifact-first output
A run produces files you can inspect, not just a screen you must trust.
Every run creates tangible artifacts: cropped files, baseline-aligned files, marker JSON, telemetry CSV, plots, reports, debug snapshots and review-ready outputs. This matters because the analysis does not disappear into the interface; it leaves evidence that can be audited, shared and inspected after the session.
CSV / JSON
Telemetry, markers and debug evidence are retained as files.
Plots
Technical plot outputs support deeper inspection outside the GUI.
Reports
Coach, Master and Muscle reports translate findings into readable form.
Review surfaces
The interface presents generated outputs rather than inventing them visually.
Baseline-relative analysis
Different measurements use the correct comparison authority.
Kineologist does not treat every aligned value as interchangeable. Timing and event order use the native baseline. Phase-matched pose and geometry use the aligned baseline. Speed and acceleration retain native magnitude. Every report difference states its source and comparison authority.
Native timing
Frames, seconds and event order.
Aligned geometry
Pose, angle and phase-matched position.
Native magnitude
Velocity, acceleration and muscle-proxy amplitude.
Traceable reporting
Every difference states its source and comparison authority.
Telemetry and TrackView
Amplitude-preserving TrackView separates timing from intensity.
Phase-locked TrackView changes where the reference curve sits in time while retaining its original physical amplitude. This avoids making a compressed phase look artificially faster or a stretched phase look artificially slower.
Horizontal phase alignment
Equivalent movement events line up in time.
Preserved magnitude
Velocity, acceleration and proxy amplitude retain their original scale.
Original authority
Native timing remains available in the Original view.
Clear report contract
Timing and intensity are reported under separate authorities.
Motion-derived proxies
Muscle proxy sequencing is derived from movement, not direct activation measurement.
Muscle proxy groups are inferred from relative joint rotations, joint motion and velocities. The system uses these motion-derived signals to visualise sequencing, contribution and baseline-relative differences, and in selected workflows they can also refine detector outputs after the first event pass. This can support movement review, but it is not EMG and must not be described as direct measured activation.
Joint rotation and velocity inference
Proxy activity is inferred from relative joint rotations and movement velocities.
Full-body muscle mapping
Motion-derived proxy channels map across major body regions.
Flexion and extension groups
Flexion and extension families make contribution patterns easier to inspect.
Muscle activity HUD
Proxy percentages can be reviewed beside the 3D motion display.
Sequencing and contribution reports
Dedicated report surfaces summarise ordering, contribution and comparison patterns.
Detector refinement
Selected detector passes can use muscle proxy evidence as a post-process to sharpen event timing.
Clear boundary
The system does not claim EMG or direct muscle activation.
Detector refinement
The muscle system can improve detector quality as well as report sequencing.
The muscle analysis layer is not only a report surface. Because its motion-derived channels expose contribution timing, the backend can use those channels as a second-pass refinement tool after an initial detector has found the likely event window. In GAIT, for example, positional foot trace and speed establish IC/TO candidates first, then muscle proxy evidence can refine those frames into sharper walking-event anchors.
First-pass detector
Position, speed or rotation signals find the first candidate window.
Muscle proxy post-process
Motion-derived contribution channels refine event timing after that first pass.
Higher-precision anchors
Refined events become stronger timing references for crops, telemetry and reports.
Still deterministic
The refined output remains an auditable artifact path, not a hidden visual guess.
Rule-based analysis
The rule layer turns research into deterministic review language.
Kineologist's core technology is not just 3D playback. Deep biomechanical research is converted into movement phases, telemetry definitions, mathematical conditions, tunable thresholds and deterministic rule logic. After telemetry is generated, the rule layer grades persistent patterns by severity and uses that severity to select safe, context-specific language from the report lexicon. With the same inputs, settings and artifacts, the system should produce the same outputs.
Research to code
Biomechanical research becomes explicit math, logic and deterministic software rules.
Telemetry inputs
Rules consume generated channels after markers, baselines and movement windows exist.
Tunable severity
Thresholds allow expert users to tune sensitivity and severity bands for context.
Lexicon selection
Severity grades drive safe Coach, Master and specialist report language.
Context-aware
Rules run inside PITCH, GAIT or deployment-specific phase/event windows.
Evidence trace
Reports remain tied to markers, telemetry, debug outputs and CSV artifacts.
Reports and multilingual operation
One evidence authority. Two report depths. Five languages.
Coach and Master reports are generated from the same evidence profiles. The Coach report prioritises practical interpretation. The Master report exposes deeper values, provenance and measurement authority. A language-neutral semantic renderer preserves the same report roles across English, Spanish, Italian, Japanese and Traditional Chinese.
Shared evidence profile
Coach and Master cannot silently disagree about the motion.
Semantic report structure
Headlines, evidence blocks and tables keep the same role in every language.
CJK-safe rendering
Japanese and Traditional Chinese use language-aware wrapping and typography.
Evidence quality and boundaries
The report states how strongly the capture supports interpretation and what it cannot prove.
Plots and technical artifacts
Generated plots provide a technical inspection layer familiar to biomechanics workflows.
Plots support deeper interpretation of telemetry channels, marker timing and movement curves outside the main GUI. They are useful for technical review, documentation, validation conversations and audit trails where a reader wants to see the shape of the underlying data.
Technical review
Plots expose the shape of telemetry rather than only the written summary.
Channel curves
Selected channels can be inspected for peaks, timing, ramps and magnitude.
Marker context
Phase or event markers help align visual curves with movement structure.
Report support
Plots provide supporting evidence for report interpretation and audit trails.
Hardware agnostic
Auto Remap helps different skeleton definitions enter the same pipeline.
Different mocap systems use different skeleton naming conventions. Auto Remap inspects the skeleton definition, extracts semantic body roles and routes those roles into the product pipeline. This reduces manual setup and makes the system more practical across different capture sources, while still needing validation across new skeleton families.
Skeleton inspection
The file is inspected for meaningful body roles.
Semantic mapping
Names are mapped into the product pipeline.
No manual config
The goal is less user setup for supported skeleton families.
Validation path
New skeleton families still need real-world validation before being treated as proven.
Lexicon Editor
Total editable language over locked analytical evidence.
Kineologist substantially expands Lexicon authority across headlines, interpretations, labels, evidence templates, action plans, review priorities, section headings and claim boundaries. Changing language does not change detectors, rule logic, thresholds, severity or numerical evidence.
Broad report coverage
Adapt substantially all human-facing report language.
Coach and Master
Both depths remain attached to the same evidence profile.
Three-row authoring UX
Editing surfaces, search, filters and state toggles stay separated.
Locked analysis
Detectors, rules, thresholds, severity and numerical evidence remain unchanged.
Threshold tuning
Expert users can tune the rule layer instead of accepting one universal judgement.
Threshold controls allow product teams, coaches, facilities or validation partners to tune overlays, rule sensitivity, marker display and severity bands. This supports expert control rather than forcing one universal judgement model across every population, method or deployment context.
Visible controls
Sensitivity, marker overlays and severity bands can be inspected instead of hidden.
Validation support
Facilities can tune and test thresholds during beta/functional validation.
Product-specific
Custom deployments can define thresholds around their own protocol.
Repeatable settings
Settings can be kept stable so repeated runs remain comparable.
Honest boundaries
Beta software, validation pathway, clear limits.
Kineologist is currently in Beta and undergoing testing at facilities and educational institutions as a pathway to functional validation. It is a movement-analysis and review tool, not a diagnostic or clinical decision system.
Beta validation pathway
Facilities, coaches, educators and analysts can help test workflows, compare outputs and shape functional validation.
Motion-derived proxies
Muscle activity views are inferred from motion, joint rotations and velocities. They are not EMG and not direct measured activation.
Readable review, not diagnosis
Reports translate deterministic analysis into human-readable review surfaces without making clinical diagnostic claims.
Next step
Want to test the engine against your movement data?
Book a demo, discuss a beta validation pathway, or explore a custom deployment built around your movement protocol.