WK02. Can we rely on engineers to save the world?

---

“We are the first generation to feel the impact of climate change and the last generation that can do something about it” Jay Inslee, Governor of the State of Washington

On the 20th of February, I attended a talk at Imperial College given by Tim Chapman, Director, of Boston Consulting Group, and Fellow of the Royal Academy of Engineering. The lecture made clear the big problem that we are facing globally with climate change and sustainability. He was mainly addressing engineers yet I consider it was a talk that can be trans-disciplinary. The main take-aways from the talk were:

• There aren’t many engineers in government positions. This means that those who are currently in power do not have a diverse skillset that truly advises the climate change problem. This got me wondering whether all education programs would benefit from having some kind of politics and economics content embedded in the curriculum, after all, so much research has been done around climate change, yet policy is playing catch-up instead of advising and applying effective legislation.

• The concept of ‘Net Zero’ in the collective aim for reducing emissions by 2030 is in itself problematic. Everything that we produce and manufacture has a Carbon cost, and therefore the use of words should be carefully considered.

• There is a need for Carbon literacy. People, even scholars, do not understand the measures by which Carbon emissions are represented. There is a need to use understandable units or metaphors if needed, to understand the real amount we are producing. CO2 can be measured in tonnes, MegaTonnes, Grams, and part per million, yet it is a lot more relatable to compare that unit to some known quantities, such as ‘the size of a football pitch’.

• Sustainability should be at the forefront of any project development, and should not be an afterthought: “Doing the right thing rather than doing the wrong thing well”

---

Initial Research Material

I made a quick search on different tools that already exist and that apply object detection/ computer vision to optimize performance. I also found some examples of ‘smart bins’ and how a rewards system is being utilized to encourage people to use them. The gathering of this material helped me to gain an initial understanding of how the technology is currently being used, and in which contexts is being implemented. The examples I have initially chosen describe the collection of geographic information system data, inventory management, collective contribution of data for research, and reward systems. These different focuses, if joined together, could see the completion of a compelling product.

---

Supervision Feedback

This week I had my first Supervision in which I presented my initial Branstorm. I have explained the different ideas in which I could focus yet I have made clear my final choice. I have been advised to prepare the following for the next session:

• Define my main problem and find academic research that can support my project idea (point 1 of Project Report)
• Reflect on why do I think my project proposal is a feasible solution to the problem
• Meet with technicians to understand the feasibility of finishing this project within the estimated timeframe
• Do research on other projects that might have used similar solutions (point 2 of Project Report)

Technicians Advice

I presented my initial project idea to two technicias that might have the required knowledge/ skillsets to advice me in the right direction. I pitched them my idea and asked them their opinion regarding my expectations and ambitions for the project. Both of them agreed on the fact that the project is ambitious yet doable. They have suggested:

• Look at examples of projects that are using React Native (UI software framework) in combination with computer vision models for object detection/ classification.
• Prepare possible plans I can diverge to in case the development of the tool takes longer.
• Use pre-existing computer vision models (YOLO, COCO, Taco) and test the level of accuracy with the data I want to evaluate.
• Undestand that I can use a pre-existing model that can be further improved with new data. Therefore, the process can be divided between two phases: 1. using pre-existing models and assessing accuracy 2. Improving the model with the addition of new data.