Introduce your group and your group mates.
Please note: The default structure for the blog is for just reference.
Feel free to change or add on if you have any other title/sections that you want to include.
For online blog guides, please visit: https://blogs.ntu.edu.sg/online-guides/.
Members of Project Ouroboros:
Ethan Chng BIE
Ethan Goh MSE
Gregory Khoo MAE
Tay Ming Sheng MSE
Li Xingchen MSE
Yuankun Zhao EEE
This blog will showcase how we implemented the Design Thinking process for each of our subsystems. In particular, how the different steps shown below were utilized in the process of coming up with the solutions that can be seen in our final product.
Empathise > Define > Ideate > Prototype > Test.
At the start, there were many ideas bouncing around at first. This is usually one of the most fun parts in a project, since its essentially an opportunity where everyone gets their chance to throw their most absurd ideas and see what sticks at the end of the day.
Initially, the original mission statement revolved around fire fighting, so there were numerous ideas, ranging from smart fire alarm systems that took inspiration from current designs and could more accurately pinpoint hazard sources, hybrid drones which could transform into a ground form and traverse under tall obstacles and venture further compared to traditional drones, and even robot dogs.
But diving deeper into our research, we were soon faced with 2 realities, which was that our ideas required time and engineering knowledge beyond what was humanly possible in 3 months, and that many of the ideas we wanted to implement would involve overhauling entire systems, which increased the complexity of the project beyond what any of us were remotely prepared for.
With this in mind, after much discussion, we decided to build upon an existing project, which was the bioinspired snake project first made by Will Donaldson. This decision was hugely based upon the fact that building on an existing project could allow us to redirect our resources to improving certain functionalities of the existing robot snake Planning ahead with a greater variety of movements in mind for the future, we went ahead with the 2D snake design since it harbored greater potential for different movement sets which would be valuable in navigating debris normally found at the scene of a disaster, where such terrain would usually be extremely uneven and unpredictable in nature, favoring flexibility in our design.
https://www.instructables.com/Bioinspired-Robotic-Snake/
The 2D snake design
used in the original design was fully tethered, had 11 segments in total, and functioned using a total
of 10 servo motors arranged in the manner as shown.
Moving ahead with this plan at the start, we started coming up with numerous ambitious features that we felt could potentially bring even greater utility to a snake robot which we wanted to further modify, while keeping in mind its intended purpose being able to help identify and reach victims trapped deep under debris.
Once again, the process of coming up with a new set of ideas swept through the team. At the time, the ideas we had were::
- Payload delivery (food packs lining the length of the snake, delivery tube for necessities, or maybe even a initial tunnel tube that would deploy ahead of the main robot snake which would then follow through a path established by the tunnel tube?)
- Mobility solutions(tracks lining the side of the snake at the front at back, extendable wheels at the head which would mimic a landing gear of an airplane)
- Wirelessly control it with a PS4 controller, make it such that we have a servo motor dedicated towards allowing it to free-look around itself
- Night vision to see victims in the dark using a repurposed home security camera
Though we had a variety of ideas at the time, we soon pivoted our focus towards mobility solutions after gaining further insights from the interviews we had conducted with HTX, professor and PhD students.
It was a given that there was the potential for encountering many unexpected obstacles, due to our unfamiliarity with executing such a concept with the additional need for further upgrades that were not taken into consideration in the original design, which did eventually come true when we faced issues during the prototyping stage.
And despite all that, it was nice to see that some of our initial ideas were still able to make it into the final design, though in a slightly different form.
We eventually settled on a roadmap as indicated below:
- Phase 1: Base snake Cabled ESP32, PCA9685, Buck Converter wired to a battery off-board the main robot body.
- Phase 2: Controllability WiFi/Bluetooth
- Phase 3: Sensors (Hazard detection, Ocular functionality)
- Phase 4: Movement solutions
- Phase 5: Payload (modular system which can be easily retrofitted onto the snake)