1 · What is VectorCam?

VectorCam combines an Android mobile app, a compact imaging attachment (macro lens + controlled illumination), on-device AI, digital data collection, offline-first operation, and centralized data management & integration. It supports mosquito classification — species groups, sex, abdominal status — where validated models exist, making routine surveillance more accessible, efficient, standardized and scalable.

🦟CollectionMosquitoes from traps
→
📷ImagingMacro + light
→
🤖AI IDOn-device models
→
✔ValidationQA checks
→
💾SyncWhen online
→
📊IntelligenceInsight dashboards
→
🎯ActionProgram response

2 · The surveillance challenge

Conventional entomology depends on scarce experts, microscopes, paper forms and delayed lab confirmation. Coverage is thin, data are fragmented, and decisions lag behind transmission. VectorCam shifts identification to the point of collection — guided, standardized, and digitally captured.

3–4 · How it works: hardware + app

📷 Imaging attachment

  • Compact clip-on unit with macro lens
  • Controlled illumination for consistent images
  • Designed for low-cost local manufacturing
  • Works with mid-range Android phones

📱 Mobile application

  • Guided capture & specimen workflows
  • On-device AI inference — no connection needed
  • Digital forms: site, trap, household metadata
  • Offline queue with encrypted sync

5 · AI & model validation

CapabilityStatusNotes
Species-group classificationValidated (study-specific)90%+ for several evaluated categories under specified conditions (Uganda 2023–24)
Sex determinationValidated (study-specific)Within evaluated workflows; see manuscript
Abdominal statusWhere supportedOnly where validated models exist
New geographies / speciesRequires re-validationPerformance must be confirmed per setting
Critical: AI outputs support field teams — they do not replace expert verification or laboratory identification where required. Study-specific accuracy is not universal product performance.

6–8 · Data, offline & integration

✔ Quality assurance

Image checks, confidence thresholds, expert review queues, duplicate detection, audit trails.

📴 Offline-first

Full workflow without connectivity; background sync, conflict handling, low-bandwidth optimization.

🔗 Integration

Centralized management, exports/APIs toward public-health information systems, district & national reporting.

9 · Evidence snapshot

90%+Accuracy (evaluated categories)Uganda cRCT 2023–24*
17s vs 34.5sMedian processing timeVectorCam vs comparison
82SUS usability score48 VHT users
240Households, 2 districtsAdjumani & Mayuge

*Provisional website content pending final manuscript & approved reporting language. Full methods & limitations: Research & Impact.

10 · Deployment & pilots

Implementation models: research validation → pilot → active implementation with training-of-trainers, district rollout and MOH integration. Status varies by country — see Global Programs.

Discuss a VectorCam pilot

Tell us your program, geography, and surveillance goals.

Request a Demonstration →