Project Nestbox

Slack: #nestbox-general, #nestbox-hardware, #nestbox-software


In collaboration with the Swiss Ornithological Institute:

Which RFID technology to use?

LF, HF or UHF?

Available products only in LF/HF. Example here.

UHF for animal tagging is only in pre-study phase. –> no benefit for us at the moment, but looking in the future, it may be beneficial (eg. for measuring more data of bird while they are away).

A first prototype was realised using HF reader and tags. Being one of the cheapest dev kits, we chose the CR95HF resp. ST95HF from ST.

Range problems as well as unsolved hardware issues (code execution would crash after a certain number of read cycles; power cycling the board does not help restoring functionality for at least 24 hours.) led to the decision to switch to an LF-based system.

System components (v2)

Rough BOM

Reader items (volume:10) price (CHF)
electronic components & PCB 25
(optional: LoRa) 20
(optional: solar cells) 10
RFID antenna 10
photodetector 5
case 15
battery 10
10 tag rings 2.5
manufacturing time ca. 5-7 hours
TOTAL BOM excl. manufacturing ca 65.- to 95.-

State of v1 Prototype

Light Barrier Test

View functional test of the light barrier.. Passage detection works and can distinguish between in/out direction.

Status: Functional & optimized
Possible improvements: lower current consumption.
Next action: Test IR LED with higher efficiency.

RFID Reader Test (NFC)

View functional test of the NFC reader with plastic bird ring. LED turns on and blinks while RFID tag gets detected. Max. tested range: <1 cm.

Status: Functional but makes microcontroller crash after 1-2 minutes.
Possible improvements: Increase range. Avoid crashes.
Next action: Test the entire system again with lower read rates.

Power management (Vbat to 3.3 V converter)

Status: Functional. Non-optimized.
Possible improvements: More efficient component at given power consumption.
Next action: Test different component for v2.

Data storage (internal FRAM)

Status: Functional.
Possible improvements: Internal storage space is too small.
Next action: Add external (SD card) storage on v2

Real Time Clock

Status: Available, but not implemented.
Possible improvements: -
Next action: Configure microcontroller to make use of RTC.

State of v2 Prototype (Jan 4. 2018)

Component Evaluation

Status: Done. Transferred to Schematic
Possible improvements: -
Next action:


Status: Schematic done. Layout pending.
Possible improvements: Double check schematic.
Next action: Start layout.

LF antenna

Prototypes with coil antennas were tested (range ~1 cm, due to high resistivity: 12 Ohms) as well as wire-wound antennas (similar range if long enough wire is used). PCB antennas with >40 windings proofed to be the best performing version if placed horizontally below the tag.

Status: Evaluation phase completed.
Possible improvements: Better matching capacitor for optimal performance.
Next action: Evaluate in real test conditions.

Weight sensor

Proof-of-concept with standard low-cost load cell successfully tested.

Status: Evaluation phase completed.
Possible improvements: Temperature compensation.
Next action: Integration of the acquisition IC on the main PCB.

Mechanical design

We are opting for a horizontal PCB antenna fixed on the nestbox entrance's lower edge. Advantage: PCB fits inside 10x10cm board size.

Status: Concept is clear. Implementation needs to be tested.
Possible improvements: -
Next action: Add wooden enclosure & ensure waterproofing.


Here are potential other technologies to exploit.

PIR motion sensor
  • promise extremely low power consumption.
  • lets us detect proximity of an owl and turn on the entire system.
RFID tag detection (to save power) using MSP430: