Physical Operations Intelligence Platform
See the operating condition before the response window closes.
In a controlled pilot, AnyVidi brings supported anonymous movement, waiting, journeys and role-level coverage into one mapped physical context.
Configured measurements form an operating condition. While it is still active, the dashboard presents a relevant next step for manager review. Footfall is the first signal. The mapped operating context shows where waiting, journey breaks or coverage gaps form.
- Observe
- Physical context
- Operating condition
- Manager review
Camera Coverage
First confirm which condition can be observed reliably.
The operating question defines the evidence required.
The Spatial Audit evaluates camera angle, image quality, stability, lighting, occlusion, overlap and covered area for the operating question.
Suitable existing views remain in use. Repositioning or targeted additions are considered only for confirmed evidence gaps.
The condition and physical step that must be observed are defined first.
Angle, image quality, stability, lighting, occlusion, overlap and coverage are audited.
Supported observations are related to the accepted physical reference after calibration.
Each view has an explicit supported event and a limit on what can be concluded.
AnyVidi uses depth-aware computer vision to relate supported camera observations to a shared 3D spatial reference. Camera coverage determines where live activity can be observed reliably, not the boundary of the mapped site.
Standard depth-aware spatial context does not require LiDAR, but it does require suitable views and calibration.We never present a blind area as live observation.
Local Processing & Anonymous Data
Process camera images on site and move only the operational data required.
In a supported pilot deployment, camera images are processed on the customer-site AnyVidi Edge PC. Continuous raw video is not sent over the WAN for standard inference.
Anonymous operational metadata, KPIs, system-health data and approved integration traffic may synchronize when the service requires them.
AnyVidi does not rely on facial recognition, biometric identity or persistent individual profiles. The technical lifecycle of anonymous journey continuity is not described until separately confirmed.
Camera streams are processed by edge compute within the customer-site network.
- Cameras
- Local network
- AnyVidi Edge
Required operational data and approved integration outputs can be synchronized when needed. WAN use is not zero.
- AnyVidi Edge
- Spatial events & KPIs
- Synchronization when required
Supported journey continuity does not rely on facial recognition, biometric identity or a persistent individual profile; its technical lifecycle remains unresolved.
Access, retention, integration and review boundaries are explicitly defined as part of deployment.
Integrations
Keep physical observation separate from the system-of-record outcome.
AnyVidi can show physical movement, waiting, journey continuation and visible interaction in supported views. A purchase, transaction, appointment, lead, quotation, test drive, service record or revenue outcome comes from the relevant system of record. The two evidence layers may be reviewed together only when an approved and implemented system-of-record integration exists.
Entry, waiting, movement, journey continuation and visible interaction.
Recorded lead, appointment, schedule, transaction, campaign or service data, when an approved integration exists.
A purchase, transaction or other outcome supplied by the relevant system of record.
Camera observation alone does not prove a purchase, transaction, conversion or revenue outcome.
Operating conditions and action signals
Show the manager the full condition, not only the threshold.
In a controlled pilot, a configured condition is evaluated with location, duration or threshold, journey context, role-level coverage and priority. While the condition remains active, the dashboard may present a relevant next step for manager review, and supported pilot deployments can generate a live operational alert through the configured notification mechanism. An alert does not establish assignment, ownership, SLA, escalation or verified resolution.
What is happening, and how long has it been active?
Which zone, journey or role-level coverage condition is involved?
What do threshold, trend and severity indicate?
A relevant next step presented for managerial judgment.
Spatial Audit
Define the operating condition the first pilot must prove.
The Spatial Audit defines the operating question, usable camera views, covered areas and journeys, evidence gaps, pilot measurements, configured conditions, Edge design, privacy boundaries and any required system-of-record context. The pilot starts with an explicit evidence boundary and jointly accepted definitions.
Plan a Spatial AuditPlatform FAQ
Clarify pilot scope, evidence boundaries and the technical model.
The Spatial Audit confirms which question each view can answer reliably.
Cameras and 3D context
Can we use our existing cameras?
Yes, when a camera view is technically suitable for the intended measurement. We assess angle, stability, image quality, lighting, occlusion, overlap, coverage, frame rate, compression and usable scene depth.
Suitable views stay in use. Repositioning or targeted additional cameras are considered only when the audit confirms a real coverage or measurement gap.
Can every camera measure the same things?
No. Camera agnostic does not mean universally compatible. A view that supports entry counting may not support proximity, height or physical-dimension estimates.
The Spatial Audit defines the KPIs each view can support and what cannot be concluded from that view.
What does the 3D operating map add beyond a heatmap?
A heatmap summarizes observed intensity on a flat image or plan. AnyVidi can place supported anonymous density, waiting, journey and role-level team-coverage events within the site's 3D spatial reference.
This shows the entrance, corridor, floor, display or service step where a condition or physical journey drop-off forms. The heatmap remains a useful supporting density layer.
Does 3D measurement require LiDAR or a depth sensor?
No, not for a standard AnyVidi deployment. Depth-aware computer vision and spatial calibration relate technically suitable 2D camera observations to the shared 3D reference.
This is not a claim of LiDAR-equivalent or survey-grade accuracy. Every intended measurement still depends on camera suitability and calibration.
When can physical measurements such as distance and proximity be used?
Distance and proximity may be used only under suitable geometry, calibration, visibility and acceptance conditions.
Product acceptance for height and body outputs remains unresolved, so they are not presented as current accepted capabilities.
What happens in blind areas?
The mapped physical site can form the full 3D spatial model, but live observability follows suitable camera coverage. AnyVidi does not invent live movement inside uncovered or blind areas.
If an operating question depends on a confirmed gap, the audit may recommend repositioning a camera or adding targeted coverage.
Data and privacy
Where are camera images processed?
In the standard on-premise architecture, camera streams travel over the customer site's local network to edge compute and images are processed on-site.
Continuous raw video does not need to travel to the cloud for standard inference.
What is sent to the cloud, and does AnyVidi use the WAN?
Anonymous operational metadata is synchronized when the configured service requires it. Required KPIs, system-health or control data, synchronization data and approved integration outputs may also use the WAN.
AnyVidi avoids loading the WAN with continuous raw video. WAN use is not zero.
Does AnyVidi use facial recognition or persistent profiles?
No. The standard platform does not require facial recognition, biometric identity or a persistent profile across visits.
The purpose is to understand the operation, not personal identity. The technical lifecycle of anonymous journey continuity remains unresolved and is not described here.
How are privacy and data-protection boundaries defined?
Purpose, access, retention, user permissions, integrations and review boundaries are defined for each deployment.
Applicable privacy and data-protection obligations are assessed against the use case, organizational responsibilities and configured data flows. AnyVidi does not make a universal compliance or certification claim.
Measurement and action
Does AnyVidi only measure footfall?
No. Footfall is the foundation. AnyVidi can relate it to occupancy, density, dwell, waiting, queues, mapped flow, anonymous journeys, transitions, exposure and role-level team coverage where the approved views support them.
The aim is to answer an operating question about what happened after entry, not to produce an unbounded list of metrics.
What can cameras measure directly?
Supported views can describe physical behavior such as entry and exit, occupancy, dwell, queues, estimated waiting, anonymous journeys, team coverage by role, continuation from a named zone and configured physical interactions.
A camera can show what happened in the physical journey. It cannot by itself prove a purchase, revenue, appointment, lead, test drive or other recorded business outcome.
Can physical journeys be connected to sales or other recorded outcomes?
The evidence layers may be reviewed together only when an approved and implemented system-of-record integration supplies the recorded outcome.
Camera observation alone does not prove a sale, transaction, lead, appointment, test drive, service result or revenue outcome, and connected-outcome association is not presented as deployed.
How do live alerts work?
A configured condition is evaluated with its location, duration, value or threshold, trend, journey context, role-level team coverage and priority.
The dashboard may present a relevant action signal for manager review, and the configured notification mechanism can route a live operational alert in supported pilot deployments. The human response remains separate; an alert does not imply automatic assignment, ownership, SLA, escalation or verified resolution.
Does AnyVidi take action without a manager?
No. AnyVidi surfaces live conditions, alerts and recommendations for manager review. Managerial judgment remains in control.
The platform is not presented as autonomously directing employees, applying layout changes or ranking individual team members.
Can we define a custom KPI?
A customer and AnyVidi may jointly define a KPI when the operating question can be grounded in supported, camera-verifiable signals and approved system context.
AnyVidi performs the initial configuration. The Spatial Audit determines the zones, events, context and evidence boundaries needed to shape the KPI.
Chain and rollout
Can we compare locations across a chain?
Configured pilot locations, regions and equivalent time periods may be compared under aligned, customer-approved definitions using supported mapped-journey, waiting, service-pressure and role-level coverage signals.
Recorded sales, revenue, lead or service outcomes require the relevant connected system. The comparison is not an individual employee ranking.
Does the model learn each location separately?
Anonymous operational metadata may be used in the cloud to update model weights separately for each location. The model learns recurring spatial and operational patterns instead of treating every site as identical.
It may produce location-specific live conditions and manager-reviewed layout or operating recommendations. It does not apply changes automatically or merge every location into one undifferentiated model.
How long does deployment take?
Timing depends on site size, camera count, network readiness, edge infrastructure, coverage gaps and the integrations required.
The Spatial Audit defines that technical scope first. A realistic deployment plan follows from the approved pilot scope and site readiness.
What does a Spatial Audit define?
It defines the operating question, usable views, confirmed coverage gaps, the 3D spatial reference, measurable journeys, pilot measurements, configured conditions, privacy, Edge design and the evidence plan.
It also defines the first dashboard conditions and action signals, plus any required system-of-record context, before the pilot begins.