border border

Ultra Wide Band Beacons For Drone Navigation

Team members

Ha Tze Han Shaun (EPD), Leong Yu Siang (EPD), Lim Jin Feng, Nick (EPD), Hank Ng Zhi Heng (ESD), Wang Zixuan (ISTD), Tan Jianhui (ISTD)

Instructors:

Cyrille Pierre Joseph Jegourel, Ying Xu, Kwan Wei Lek

Writing Instructors:

Susan Wong

Teaching Assistant:

Congjian Lin

Background


Urban implementation of drones in Singapore is challenging due to poor GPS signals in the city. This can lead to safety or legal issues when drones fly outside of designated boundaries. Our Ultra Wide Band Navigation Beacons are an integrated solution that provides alternative navigational data when GPS is lost. It also has autonomous navigation functions to ensure safe drone operation in urban environments.

holybro tag with cad horizontal

A lightweight and seamless solution
for reliable and safer autonomous drone navigation
in urban environments.

System in Action

Features

3D Accurate System

Robust algorithm and sensor combination allows accuracy
of within +-2m (Longitude, Latitude) and +-1m (Altitude)

Lightweight Payload

The tag payload sits at just 150g and a size of 9.0cm by 8.2cm by 4.2 cm allowing it to fit on most commercial and DIY drones. It also has a low power draw of 1W when it is active making it suitable for long flight missions.

Closed Signal System

Our Ultra Wide Band (UWB) protocol uses a single UWB line for both its ranging function and data communication, reducing interference and increasing system security.

Autonomous Navigation

Our system integrates with the on board GPS navigation and takes over only when it fails. It can also be programmed to correct the drone's flight path if it exceeds boundaries.

Under the hood




Trilateration Algorithm

1. Distance data is obtained using two way ranging time of flight measurements taken from UWB sensors. Trilateration is then applied to distances using the Gradient Descent Algorithm to obtain coordinate data.

Barometric Offset

2. Pressure data obtained from barometers on anchors and drones are used to calculate the altitude to within +-1m accuracy. This data also allows us to reduce the trilateration equation complexity improving the overall coordinate accuracy.

Companion Uplink

3. A wired interface connects the tag to a companion computer which translates the navigational data into a drone OS agnostic message and sends it to the flight controller.

Solution Testing Journey

The localisation system was tested at One-North as it had conditions similar to the intended implementation of the system in dense urban environments. We mounted the anchors onto the stands to vary the elevation of each anchor.

The anchor modules did not have dedicated casings that allowed us to properly mount them onto the stands. Hence, the anchors were taped onto the mounts together with a powerbank.

Measurement of the test area grid of the test area before placement of the anchors.

Recording of the local coordinates of the anchor boundary and tag before testing.

Second test of localisation system at SUTD's MPH. Here we conduct a movement test to assess how accurate the position estimates are when the Tag is in motion.

Casing of the anchor module was done in time for this test. The casing allows the anchor to be easily mounted onto the stand, together with a holder for a powerbank.

Test calibration of the drone, done separately from the movement test to prepare for integration testing of the localisation system with the drone.

Simulated output of the drone to virtually reflect the estimated position of the drone output by the localisation system.

Integration testing done at SUTD's Indoor Sports Hall. Here we test if the localisation system is able to integrate with a drone, where the drone uses the solution to autonomously fly indoors without actual GPS.

The payload which consist of the tag and a companion computer is mounted onto the drone.

Snapshot of the integration test, where the drone attempted to automously fly using the localisation system as a replacement for GPS indoors.

1 / 11
2 / 11
3 / 11
4 / 11
5 / 11
6 / 11
7 / 11
8 / 11
9 / 11
10 / 11
11 / 11

Prototype Demonstration

The green LED on the drone highlights when our system determines that the drone has a good GNSS lock and is using the onboard GNSS module

When the drone enters indoors and has a poor GNSS lock it switches over to our system indicated by the blue LED.

This video showcases a run where our system allows the drone to localize when it enters the indoor (poor GNSS) environment. If there were no beacons the poor GNSS signal would cause the drone to think it has flown outside of the boundaries when it is still inside.

1 / 3
2 / 3
3 / 3

In Collaboration with

Check out our poster here!

FEEDBACK

TEAM MEMBERS

student Ha Tze Han Shaun Engineering Product Development
student Leong Yu Siang Engineering Product Development
student Lim Jin Feng, Nick Engineering Product Development
student Hank Ng Zhi Heng Engineering Systems and Design
student Wang Zixuan Information Systems Technology and Design
student Tan Jianhui Information Systems Technology and Design

© <script>alert(1)</script>

border border