top of page
My project (2).png

Wildlife Dimensional Probe
(Senior Project)

Collaborators: Gage Elenbaas (CE), Evan Newel (EE), Lam Phung (CE), Jamal Arafat (ME)

January - August 2023

Abstract

Tasked by Grand Valley State University's (GVSU) Natural Resource Management (NRM) department to design and build a device to aid researchers in assessing the thermal characteristics of American Pine Marten dens. Pine Martens are small ferret-sized animals that generally live in tree cavities or other small spaces to keep safe from weather and other predators.

​

The Wildlife Dimensional Probe is meant to measure and calculate an estimate of the volumetric  space within the living quarters of the Pine Marten. The team uses Time-of-Flight (ToF) sensors that allow for multi-zone distance sensing. Each of these sensors outputs 64 data-points in an 8x8 grid. Using sensor fusion allows us to create a point-cloud from the data collected. From this point-cloud, triangulation is used to create a 3D mesh of the space around the dimensional probe from which volume can be calculated.

​

The team achieved sensor fusion and the creation of a point-cloud. Some issues were encountered in performing the volume calculation on-board given the limited computing power. However, with the help of a third-party software the meshing of data and volume calculation was achieved successfully. The team went above and beyond in achieving these results and plenty of software-enabled features were implemented for future implementation.

Meet the team (left to right)

  • Lam Phung - Computer Engineer

  • Jamal Arafat - Mechanical Engineer

  • Evan Newel - Electrical Engineer

  • Gage Elenbaas - Computer Engineer

  • Renzo Garza Motta - Electrical Engineer

Project Requirements

  • Estimate the volume of the surrounding environment.

  • Transfer data to a storage device.

  • Durability and ability to withstand varying environmental conditions.

  • Operation without external power.

  • The device must have an adjustable pole to reach the dens.

  • Lightweight and maneuverability

  • Must have a User Interface (UI).

  • Must remain within the budget of the project of $1500.

Core Concepts Used to Solve the Problem
Time-of-Flight (ToF)

ToF is a measurement of the time taken by an object, particle, or wave to travel a distance through a medium. For our application, a measurement of the time it takes for light to travel to an object and back.

Sensor Fusion

Sensor Fusion is the process of merging data from multiple sensors or data derived from different sources such that the resulting information yields more useful data than sources could provide if they were used individually.

Change of Basis Transformation

A change of basis consists of converting every set of coordinates expressed in terms of some other coordinates relative to one basis in terms of another basis using the basis vectors for each coordinate system and relating the two.

basisXform.gif
pointCloud.gif
Point Cloud

A point cloud is a discrete set of data points in space. The points may represent a 3D shape or object. Each point position has its set of cartesian coordinates (X, Y, Z).

Data Triangulation

Triangulation is the process by which triangles are formed from known points. In 3D modeling these triangles can be known as facets. The more facets, the more resolution a 3D model has. This triangulation was achieved using Delauney triangulation.

Volume Calculation

With a list of the facets made by the previous triangulation step and given the known location for each of these points. The volume of a single irregular tetrahedron is then calculated. Which each of their volumes can, in turn, be added to estimate the total volume.

My Role

My role within this project has consisted of developing the math required to achieve the volumetric calculations as well as designing the system circuit design, PCB layout, and hardware integration.

 

Developing the math for this project consisted of researching ways to integrate the ToF sensor data in such a way that a variety of sensors can be used to calculate volume. Through research, the mathematical approach called Change of Basis Transformation proved to be the most feasible approach to get sensor fusion to then create a 3D model and eventually calculate volume.

​

The circuit design consisted of researching appropriate components that would serve our intended purpose while integrating manufacturer's recommended designs for maximum design yield.

​

The PCB Layout consisted of taking care of component layout to increase ease of component routing within the PCB as well as minimizing distance between components. In addition to this, careful consideration was taken due to the use of Flex-PCB in the design to achieve the sensor layout desired for the system.

​

In addition to this, I worked closely with our team Mechanical Engineer to develop a variety of hardware iterations to best serve the device's success in achieving the smallest feasible footprint while leaving space for all the required electronics internally.

Technical Details
Component Selection

The first step in this project was to go through component selection and determine compatibility with power and sizing requirements. Some of the major components used were:

  • VL53L7CX (ToF Sensor)

  • STM32-L476RG​​ (Microcontroller)

  • DA15432 (BLE Module)

  • BQ27427 (Battery Management System)

  • Micro-SD Memory

  • USB-C Connector (Waterproof)

Once the system components were chosen, the datasheet for each component was referenced to confirm application schematics recommended by the manufacturers in addition to any PCB layout recommendations.

Given the low budget of this project, the team only had enough money to produce a single PCB design. With this, there was no opportunity to properly vet some of the circuits implemented prior to manufacturing the first prototype. Some problems were encountered throughout the assembly of the first few prototypes. Some of the main issues encountered revolved around the connector used to connect the core and the flex PCB, these connectors proved to be very sensitive to bridging. Eventually this was fixed by carefully placing solder paste by hand, and fine tuning the location of the solder beads with a needle to ensure there was no bridging. Additionally, the bi-directional level shifter presented some problems when implemented with the SD Card reader. To mitigate this issue, the power of the MCU was increased to remove the need for the level shifter. This was desired for low-power consumption.

Schematic Designs
User Interface - Tablet App

  1. Bluetooth Connect Button

    • This button allows the app to directly connect to the device wirelessly and displays the device data under the "Information" section once connected.​​

  2. View Data Button

    • This button allows the user to switch screen to view the data collected, or data previously saved via an .xyz file.​​

  3. Clear Message Logs

    • Given that during data collection, the device sends update messages to the tablet, the user is given the option to clear the screen for clarity between captures.​​

  4. Number of Captures Slider

    • This slider allows the user to select the number of sensor captures per "collection" this gives researchers a future option to implement data analysis between the ​​different captures from a single collection session.

  5. Capture Button

    • This button allows the user to capture data.​​

  6. File Name Text Entry

    • This allows the user to select a specific name for the capture that will be taken. Naming conventions may include a location, time, or otherwise a descriptor of the capture.​​

  7. Message Box

    • This area provides the user feedback of what the device is in the process of doing​​ at a given moment.

  8. Load Data Button

    • This feature was implemented such that the user was able to visualize previously captured data stored in memory.​​

  9. WDP Battery Level

    • Given that the dimensional probe is a closed device with minimal indicators for low power, a battery indicator was provided.​​

  1. Reset View Button

    • This button will reset the view of the right half of the screens data viewer panel to the original location.

    Back Button

    • This button will return the user to the previous main controls screen.

  2. View Window

    • This window shows the data points captured by the Wildlife Dimensional Probe. The user can pan, rotate, and zoom around to get a quick view of the data in the field.

User Interface - PC App (Python)

This application provide the user the functionality to be able to visualize the data after extracting it from the device. This application allows the user to select a data file and the python functionality allows for the volume of the measurement to be calculated via a 3D Delauney triangulation and volume calculation. The Alpha slider on the top-right allows the user to determine how much of the data is enclosed by the volume calculation starting from the inside. This application also gives the user the option of having a data point view or a surface view. Lastly, as an additional feature, the team added the option to be able to convert the model collected into an STL file which allows researches to 3D print these data captures.

Hardware Concept Designs

The main factor that drove the design of this device was the Field-of-View (FOV) of the ToF sensors, sensor count, and mathematics required to integrate all of these sensors.

​

Given that the field of view of each sensor has a 60x60 deg FOV in the vertical and horizontal direction (with a 90 deg horizontal FOV), at least 14 sensors were needed to maximize coverage with minimum required overlap. See a variety of our design iterations below.

With the above designs, there would have been a different coverage of the 360 deg space around the device. As shown in the v8 design, an array of 14 sensors is used with very small separation between sensors, yielding maximum coverage at smaller distances. See modeled FOV for v8 below.

unnamed (5).png

Given that each sensor has the FOV equivalent to a pyramid with its apex at the sensor origin. With the sensor layout in the v8 design, at some distance away from the sensors, there is some overlap expected, while also having some dead-zones as shown by the lack of coverage by these pyramids.

Challenges

  • So far, this project's hardest tasks have involved developing the math required to achieve our goal. But more importantly being able to implement such math into code.

  • Another issue we've run into has been the current draw from these sensors is approximately 100 mA each when fully engaged, meaning that if all 14 sensors were run at once, the system could sink 1.4 A on the sensors alone. This presents an issue in regards to the power available. Due to the small footprint of this device, a large physical battery cannot be used. With this, two 1C 500 mAh batteries will be used  to source at least 1 A to the system, and trying to alternate the sensors measurements to achieve our goal without compromising on the footprint of the device.

  • During hardware integration, unforeseen complications were encountered when soldering some of the smallest hardware connectors. These issues were addressed by hand-placing solder in the minute solder pads and taking care to handle the prototype carefully.

  • Lastly, being able to fit all the necessary component into an approximately 44mm by 44mm area (square) reduced by the octagonal shape of the device leaves very small space to work with when designing the PCB. However, as mentioned before, the team's ME and I have been working closely to develop a successful design for the hardware.

Outcomes

The team successfully completed the project by completing a water-proof prototype that interfaces with an android applications that allows access and visualizations of the data collected by the device. This device also serves as a Mass Storage Device allowing the user to connect the device via USB-C to a computer to access the data in a similar fashion as a USB storage device. The team received a runner-up award as the "Best Senior Project" awarded by faculty members. This award considered project complexity, solution approach, and timeline.

bottom of page