An AutismAI Lab application

Research-grade movement analysis in the browser.

Volitiq turns any device with a camera into an objective motor assessment tool. Computer vision extracts skeletal pose from standard webcam video; the platform converts it into quantitative biomarkers that researchers, clinicians, and caregivers can act on. No installation, no specialist hardware, no clinic visit required.

Browser-native
No hardware barrier

Runs on any modern device with a camera. No installation, no plug-in, no specialist instrumentation. The webcam is the sensor.

Quantitative
Numbers, not impressions

Every clip produces a structured feature set: postural sway, head movement variability, gait spatiotemporal indices, bilateral asymmetry, smoothness metrics.

Cross-population
One pipeline, many cohorts

Autism-validated, with the same feature extractor generalising to fall risk, post-stroke recovery, and elderly mobility research.

The challenge

Motor differences are documented, but measuring them objectively is gated behind specialist clinics.

Motor differences appear in 34–80% of autistic individuals depending on the cohort and the measurement instrument used. They are clinically meaningful, they are reproducible across populations, and they are systematically under-measured because traditional motor assessment requires force transducers, motion capture systems, or other equipment confined to a small set of research clinics.

The same gating problem applies to aging research, post-stroke rehabilitation, and disability mobility studies. The biomarkers exist, the methods exist, but the apparatus does not travel. Volitiq removes the apparatus from the equation. The skeletal-pose features that drive modern motor research are recoverable from ordinary webcam video, with accuracy that already supports research use in many contexts.

What Volitiq measures

A structured feature set from every clip.

Postural sway

Centre-of-mass micro-oscillations recovered from standing pose, a documented digital phenotyping signal in ASD toddlers and a fall-risk indicator in older adults.

Head movement variability

Frequency and amplitude of atypical head movement patterns, validated in ASD literature as an early behavioural marker (PMC6242931).

Reach and tool-use kinematics

Joint-angle trajectories during everyday actions: reaching, grasping, manipulating objects, transferring between hands.

Gait spatiotemporal features

Step length, cadence, stride variability, double-support time, recovered from walking sequences in the camera frame.

Bilateral asymmetry indices

Left-right discrepancies in coordinated movement, relevant to autism motor signatures and to post-stroke rehabilitation tracking.

Movement smoothness metrics

Jerk-based and spectral-arc-length scores quantifying the qualitative smoothness of voluntary movement.

The feature library is extensible. Researchers needing a measurement that is not yet in the standard panel can request it through the collaboration channel; the underlying skeletal-pose stream supports many additional derivations.

Populations served

Autism is where we started. The pipeline generalises.

Autism
Primary validation cohort

Motor differences are documented in 34–80% of autistic individuals. Volitiq operationalises the digital phenotyping literature into accessible measurements that any researcher or clinic with a webcam can collect.

Aging and fall risk
Active generalisation

Gait speed and gait variability are clinically validated biomarkers of fall risk, frailty, and early cognitive decline. The same skeletal-pose pipeline that scores autism motor signals produces these gait features directly.

Post-stroke recovery
Exploratory

Bilateral asymmetry and movement-smoothness scores are standard rehabilitation outcomes. Volitiq makes them measurable between clinic visits without specialist equipment, opening the door to higher-frequency outcome tracking.

Neuromuscular and developmental disorders
Future

The kinematic feature library is population-agnostic. Any condition with a measurable motor signature is a candidate population, provided the right normative reference can be assembled with research partners.

How a session works

From recording to research-ready features in minutes.

1
Record

The participant performs a task (standing, walking, reaching, tool-use) in front of an ordinary webcam. Guided in-app instructions keep the framing consistent.

2
Extract pose

Skeletal landmarks are extracted frame by frame using validated computer vision models. The raw video can stay client-side when feasible.

3
Compute features

The kinematic feature library converts the pose stream into quantitative biomarkers: sway, gait indices, asymmetry, smoothness.

4
Export

Features land in research-ready formats with full provenance. Cohorts stay scoped to the partner research team that collected them.

Under the hood

Engineering choices that keep the barrier low.

Capture
React frontend; webcam-based, browser-native; no installation, no plug-in.
Pose extraction
MediaPipe skeletal pose, run client-side where feasible; server-side fallback for low-spec devices.
Inference
Python on AWS Lambda; horizontally scalable per-clip processing.
Storage
DynamoDB for structured features and metadata; S3 for raw and intermediate artefacts under researcher control.
Export
Research-ready data formats (CSV, JSON, parquet); per-cohort scoping; configurable retention.

What Volitiq does not claim

Three explicit non-claims.

Not a diagnostic instrument

Volitiq produces research-grade measurements. It is not a medical device and does not provide a diagnostic claim about any condition.

Not a motion-capture replacement

For sub-millimetre precision, marker-based motion capture remains the gold standard. Volitiq optimises for accessibility at research-useful accuracy, not laboratory ceiling accuracy.

Not an autonomous clinician

Every score and trend exists to inform a human reader. Researchers interpret features. Clinicians decide on care. The platform surfaces signal cleanly; the interpretation is theirs.

Roadmap

Deeper kinematics, more populations, tighter integrations.

Deeper kinematics
  • Hand and finger pose for fine-motor research.
  • Inter-segmental coordination metrics.
  • Improved low-light and occlusion robustness for in-home capture.
Cohort generalisation
  • Normative references for older-adult and post-stroke populations.
  • Task templates for standard rehabilitation assessments.
  • Pilot partnerships in geriatric mobility research.
Researcher experience
  • Cohort dashboards with group-level statistics and individual deviation scoring.
  • Direct R and Python notebook integrations for feature pulls.
  • Validated normative bands per task across age and cohort.

Try Volitiq, or bring it to your cohort.

Researchers can launch the app and capture a session in minutes. Labs and clinics interested in deploying Volitiq across a study cohort, or in proposing a new measurement, can reach the team through the collaboration page.