WSU Thesis Update

Current IntelliKeep progress from the completed project report.

This page now reflects the April 2026 Weber State University report for IntelliKeep: a complete BLE asset-tracking system with custom tags, an ESP32-S3 base station, cloud sync, browser monitoring, and mobile alerts.

Report status

Prototype-level system architecture, hardware, firmware, backend, and mobile workflows are documented and evaluated in the thesis report.

Updated from thesis report dated April 23, 2026

Main remaining gap

The strongest unfinished item is long-duration CR2032 validation under realistic Bluetooth pulse loading, plus broader field and stress testing.

Prototype Snapshot

Key measured numbers and implementation constraints pulled directly from the report.

1.5 µA

Measured deep-sleep current

Observed on the tag hardware with the low-current regulator implementation installed carefully.

200 m

Open-area BLE range

Reliable communication between the tag and base station in open-area testing.

150 ms

Advertising window

Short beacon interval used in the final low-power tag wake cycle.

270 ms

Total awake time

Steady-state wake duration before the tag returns to deep sleep.

5 min

Beacon cadence

Final low-power operating interval used by the current firmware direction.

32 tags

Current embedded whitelist ceiling

Implementation limit of the present NVS-backed local storage model on the base station.

Project Progress

The biggest changes between the earlier concept and the finished thesis prototype.

Architecture

End-to-end system defined and integrated

The report describes a complete four-layer system: BLE tags, an ESP32-S3 base station, a Flask cloud backend, and browser/mobile clients with notifications.

Hardware

Custom tag and base station hardware prototyped

Both the BLE tag and the base station now have custom PCBs, tuned NFC support, tamper sensing, and printed enclosures instead of remaining at dev-board stage.

Firmware

Tag design pivoted to a lower-risk beacon model

The implementation moved away from connection-oriented BLE and into a short beacon-only wake cycle, reducing complexity and keeping presence logic at the base station.

Validation

Prototype-level requirements largely verified

The report concludes that most core monitoring, enrollment, cloud sync, and dashboard requirements are met, with the main remaining gap centered on CR2032 validation and larger stress testing.

Path Forward

The report frames IntelliKeep as a strong prototype moving toward a more deployable product.

Complete funding and pilot-deployment preparation so the prototype can move into more formal field use.
Finish FCC, trademark, and patent work needed for a deployable commercial hardware configuration.
Deepen integrations with rental-property and distributed-site management workflows.
Develop more deployment-ready hardware variants, including weather-resistant options.

Measured Results

Recent testing emphasized power behavior, radio range, low-power firmware direction, and reliability mechanisms.

Power profile

Deep-sleep current measured at about 1.5 µA.
Wake current is staged rather than flat: about 10 mA during bring-up and about 43 mA at BLE peak.
Battery freshness is cached so full battery measurement does not need to run on every wake.

Radio performance

Open-area testing showed reliable tag-to-base communication out to roughly 200 m.
The measured range indicates the open-space radio path is not the primary limiting factor in the prototype.
The report still frames indoor deployment behavior separately from the open-area result.

Battery direction

The TPS610994YFFR power stage performed successfully with an LIR2032 rechargeable coin cell.
Standard CR2032 testing remained the major hardware constraint during Bluetooth initialization.
Long-duration CR2032 validation is the clearest remaining hardware gap in the report.

Reliability strategy

The base station uses Wi-Fi reconnection with exponential backoff and heartbeat-based cloud health monitoring.
Missing-state decisions are time-windowed so a single missed advertisement does not immediately imply loss.
RSSI-based out-of-range classification and local NVS caching keep the system useful during transient issues.

Performance Highlights

These bars summarize standout technical results discussed in the report.

BLE range versus 15 m minimum requirement 200 m observed

About 13.3x the stated minimum indoor line-of-sight requirement.

Collision Probability Model

The beacon-only design keeps estimated single-wake collision probability low for the intended 3 to 5 tag deployment range.

3 tags

0.20%

5 tags

0.40%

10 tags

0.90%

25 tags

2.37%

50 tags

4.78%

The report notes that the 11-minute missing window gives a tag roughly two chances to be observed before being classified missing.

Requirement Status

The report concludes that IntelliKeep meets the majority of its functional and performance goals at the prototype level.

Functional requirements

Satisfied 7

Substantially satisfied 3

Partially satisfied 2

Architecturally supported 1

Performance requirements

Satisfied 2

Substantially satisfied 1

Reasonably supported 2

Architecturally supported 1

Open gap 1

Prototype Gallery

Hardware, enclosure, and circuit images sourced from the thesis project materials.

Prototype hardware

Printed prototype enclosures

Physical prototype enclosures for the IntelliKeep tag and base station.

Tag PCB

Custom IntelliKeep BLE tag PCB prototype.

Base-station PCB

Custom ESP32-S3 base-station PCB prototype.

Mechanical design

Tag enclosure exterior

Compact cylindrical tag housing with snap-fit construction.

Tag enclosure interior

Internal packaging arrangement for the tag electronics.

Base enclosure exterior

Standalone base-station form factor developed in Fusion and 3D printed.

Base enclosure interior

Internal supports for PCB mounting, connector access, and antenna placement.

Key circuit blocks

Regulator stage

Ultra-low-quiescent-current power stage built around the TPS610994YFFR.

NFC enrollment block

NFC hardware used for close-range tag enrollment and identity transfer.

Tamper sensing

Photodiode plus comparator path for low-power enclosure tamper detection.

Battery sensing path

Switched divider design that avoids permanent sleep-current leakage.

Lessons Learned

The report’s main engineering lessons came from real hardware bring-up rather than only from software integration.

Coin-cell BLE hardware is constrained by transient behavior and regulator layout, not just average current math.
The switched battery-measurement path can fail in ways that initially look like firmware defects.
Tamper sensing required empirical tuning on assembled boards before it behaved reliably.
Moving from connection-oriented BLE to a beacon-only architecture simplified the system and reduced energy overhead.