Skip to main content

Product-Level Engagement Analytics via ESL-Triggered Spatial Calibration

1. Executive Summary​

Retail analytics has historically treated products as static entities bound to fixed shelves or planograms. In reality, products move, layouts evolve, and temporary displays disrupt static mappings — making traditional shelf-based engagement metrics increasingly inaccurate.

This white paper introduces a new class of product-level analytics enabled by the integration of Electronic Shelf Labels (ESL) with video-based footfall and behaviour analytics. By synchronising time-scheduled LED flashes on individual ESL tags with ceiling-mounted cameras, retailers can dynamically determine the true spatial location of products and directly associate shopper behaviour with specific products or product groups, not just shelf zones.

2. The Core Problem: Shelf ≠ Product​

2.1 Limitations of Traditional Product Analytics​

Most retail analytics systems assume:

  • Products are fixed to predefined shelf coordinates
  • Engagement occurs at a “zone” or “fixture” level
  • Planograms are static and reliable

In practice:

  • Products are frequently relocated (promotions, shortages, replenishment)
  • End-caps, dump bins, and pop-ups break planogram logic
  • Shopper interaction is with products, not shelves

This results in:

  • Misattributed engagement
  • Inaccurate dwell-to-product correlations
  • Poor visibility into true product performance

3. Concept Overview: ESL as a Spatial Anchor​

3.1 Key Idea​

By leveraging ESLs with built-in LED indicators, FootfallCam introduces a time-synchronised spatial calibration mechanism:

  1. FootfallCam V9 schedules a flash event for a specific ESL tag (by serial number)
  2. A MQTT command is sent via SDK to the ESL system
  3. The ESL flashes its LED at a precise, scheduled time
  4. Pro1 / Pro2 ceiling cameras detect the flash in video frames
  5. Computer vision localises the flash position in 3D space
  6. The ESL’s physical location is dynamically mapped and stored
  7. Shopper trajectories and dwell are now associated with that product

This converts ESLs into active, camera-verifiable location markers.

4. System Architecture​

4.1 High-Level Flow​

V9 Scheduler
    ↓ (MQTT / SDK)
ESL Control System
    ↓ (LED Flash @ T)
ESL Tag (LED On)
    ↓ (Video Capture)
FootfallCam Pro1 / Pro2
    ↓ (CV Detection)
Spatial Mapping Engine
    ↓
Product-Level Analytics (VOM)

5. Technical Requirements​

5.1 ESL Capabilities (Mandatory)​

The ESL system must support:

  • Individually addressable ESL tag IDs / serial numbers
  • Built-in LED with sufficient brightness for ceiling-mounted detection
  • SDK or API access (MQTT preferred)
  • Time-synchronised execution of flash commands
  • On-demand and scheduled LED control

5.2 FootfallCam Capabilities​

  • V9 Scheduler
  • Define flash schedules (single or batch)
  • Manage retry and validation logic
  • Camera Analytics
  • LED flash detection (temporal + luminance filtering)
  • Spatial triangulation within camera FOV
  • Cross-Validation
  • Confirm flash detection against scheduled time
  • Reject false positives
  • Dynamic Mapping
  • Store ESL → physical coordinate → product linkage

6. Accuracy Considerations​

6.1 Environmental Factors​

  • Camera angle and height
  • ESL LED brightness and diffusion
  • Ambient lighting conditions
  • Distance from camera to ESL

6.2 Mitigation Strategies​

  • Multiple flash cycles for confirmation
  • Flash patterns (e.g. pulse sequences)
  • Confidence scoring per ESL localisation
  • Periodic re-calibration schedules

7. From Shelf Analytics to Product Engagement Analytics​

7.1 New Analytics Dimensions Unlocked​

With dynamic ESL localisation:

  • Engagement is tied to product identity, not shelf ID
  • Shopper dwell can be attributed to:
  • Specific SKUs
  • Product groups
  • Promotional displays
  • Products can be tracked even after relocation

7.2 Behavioural Insights Enabled​

  • True product-level dwell and conversion proxies
  • Shopper flow relative to moving products
  • Promotion effectiveness independent of placement
  • VOM (Visitor Operating Model) by product cluster

8. Business Value​

8.1 For Retail Operations​

  • Accurate analytics without manual planogram updates
  • Reduced reliance on staff audits
  • Real-time validation of product placement

8.2 For Merchandising & Marketing​

  • Measure engagement impact of relocation
  • Compare identical products across locations
  • Optimise temporary and promotional displays

8.3 Strategic Differentiation​

This approach creates a defensible analytics layer:

  • ESL vendors provide price & ID
  • Cameras provide people flow
  • FootfallCam connects product, space, and behaviour

9. Deployment Models​

  • Pilot: Single aisle / category
  • Periodic calibration (off-peak hours)
  • Event-based recalibration (after major layout changes)
  • Scalable across hundreds or thousands of ESLs

10. Conclusion​

By synchronising ESL LED flashes with video analytics, FootfallCam enables dynamic, product-accurate engagement measurement in real retail environments where products move and layouts change.

This integration represents a shift from static shelf analytics to living product intelligence, unlocking a new dimensionality of insight for modern retailers.