The frisbee perspective🥏: Gliding into research one disc at a time
🚨DISCLAIMER 🚨
I am not a world renowned researcher/scientist and the views that I present in this article are my own.
Observing the rapid advancements in the field of Artificial Intelligence (AI), a question that naturally comes to an aspiring researcher’s mind is how can I incorporate these techniques into my research. But a better question to ask might be, do I really need AI to solve my problem? With similar questions and assumptions in mind, I started my journey into research through my master’s thesis. Along the way, I understood what a research problem is, how to formulate one and how to successfully solve it. This article is an attempt on explaining the process from my standpoint.
Let’s start by selecting our research problem from the field of disc throwing (frisbee). Why disc throwing? The game encompasses a simplistic design with scope for improvement and can be easily understood by a wide community of researchers.
Next, let’s define our research objective:
To improve the game of frisbee.
Improving the game entails enhancing the Quality of Experience (QoE) for the players involved. Consider the current version of the game as the baseline version that can be treated similar to a standards specification document.
Before we begin our search for ideas to accomplish our objective, let’s quickly summarize the game: frisbee 101: frisbee is a flying disc with an airfoil shape. Let’s assume that Alice and Bob are the only players. If Alice throws the disc and Bob grabs it (without dropping it), then it’s considered as a success by both the players. Several factors ranging from player height, distance between them, weather conditions etc. affect the QoE but for simplicity let’s assume that the players are almost identical and equally skilled but there is still an uncertain medium/channel between them (e.g. high speed winds) which is responsible for the distortions along the path of the disc from source towards its destination.
⭐Take away lesson: List down assumptions at the beginning and keep them pragmatic. Start by solving a simpler version of the problem by relaxing some of the constraints involved.
Following are some of the potential solutions:
- Changing Properties Approach (CPA): Utilizing different materials for the disc. In order to implement this, we have to simulate the properties of the material in a lab, create prototypes to analyze behavior of the disc, perform wind tunnel testing etc. to get an improved disc.
- Changing Environment Approach (CEA): Creating a controlled environment for playing the game like an indoor arena. In order to implement this, we have to assemble a standards committee that decides on the appropriate dimension and facilities of the arena. In contrast to the previous approach it would require longer time as well as significant amount of resources to achieve this, also it would depreciate the appeal of the game as it wouldn’t be accessible to all.
- Changing Rules Approach (CRA): Formulating a new set of rules for playing the game. In order to implement this approach, we would require to form a committee of subject matter experts and top players who then formulate new rules and guidelines. However, in the real world, most of the research challenges are governed by rules primarily based on underlying physics and mathematics (and are thus immutable).
- All In Approach (AIA): Incorporating sensors, and thrusters onto the disc. In order to implement this approach, sensors calculate the wind direction and speed, using this data, computation is performed on the disc or at the edge server using State-of-the-Art (SOTA) AI models to predict and add the thrust needed. This method provides an accurate path for the disc but it might take a long time to compute and implement the necessary corrections. It is also plausible that the thrusters increase the weight of the disc to the point where it doesn’t even glide.😟
⭐Take away lesson: Create your own acronym.
So isn’t there something we can do that won’t consume extra resources and development time but would still provide a significant increase in the QoE for the players?Any ideas?🤔
The answer is YES! 🎉 and one of the possible solution💡: Strap a LED light🚨to the top of the disc (a.k.a LED moment).
This small improvement while being inexpensive, adds a tiny weight to the disc but improves the overall QoE by revealing a new playing dimension. Now, the game can also be enjoyed in the dark which was not previously feasible. In a similar manner, research, I think should uncover a small but significant improvement (with or without utilizing the SOTA).
