Blog 01

Hello everyone, I'm Jiayu. This is my first blog!

About me

As you can see from the photos below, I really enjoy traveling and trying the local cuisine in different places. As Xu Wei (A singer in China) sings:
Life is more than just what's in front of us — there's more out there.

The following photos represent my most challenging trip. It was to Xizang  (located in the northwest of China. The altitude there is very high, meaning the air is thin, and it is very easy to experience altitude sickness (lack of oxygen leading to dizziness, headache, nausea, and in severe cases, even endangering life safety). Any strenuous exercise would make you short of breath. So when at a high altitude, my two friends and I could only move forward like old men. After a long journey, we finally reached the Qomolangma Base Camp (at an altitude of 5,200 meters). Although it was very painful at that time, looking back now, I feel that everything was worth it. I call this trip "viewing the scenery with my life".

I think traveling not only brings me joy, but also makes me realize that there are such sceneries, cultures and people in this world, living their lives in such different ways.

Interesting post from previous blogs

This steampunk bot immediately caught my attention. The design is very unique and the appearance is extremely cool, as if it came from a game set in the Steam Age. When combined with the heartbeat sound, the "heater" on the top of the head also starts to "heat up". Besides its initial function, it also contains many other features, such as real-time detection of heart rate, pulse, steps, and so on.

Extra cool stuff with bin bot

Building upon binbot's ability to identify trash (determining whether it can be disposed of in the bin), a "trash-picking" function could be added (as shown in the AI-generated image below). My idea is to combine bin bots and robot vacuums (it's a bit silly to simply have the big bin bots sweep the floor). Different bin bots could hold different types of trash (e.g., recyclable and non-recyclable), thus solving the trash sorting problem. This would also eliminate the need to manually empty the robot vacuum, and it could be placed at home, saving time on housework.

About State Machine Diagrams

History:

The concept of the Finite State Machine (FSM) emerged from early research in automata theory and computational models of discrete systems. Initial ideas can be traced to the logical neuron models of Warren McCulloch and Walter Pitts in the 1940s. During the 1950s, FSMs were formally defined through the work of Stephen Kleene, George H. Mealy, and Edward F. Moore. The theory was consolidated by Michael O. Rabin and Dana Scott in 1959, establishing FSMs as a foundational model in computer science and engineering.

My Drawio Diagram:

Cool example project that used state machines diagram:

Next, I will briefly introduce finite state machines in game.

Why do games need state machines?

Early video games had very limited CPUs and memory. Developers needed AI systems with the following characteristics:

• Determinism

• Lightweight

• Easy to debug

• Predictable to the player

This led to the widespread adoption of finite state machines (FSMs) and their becoming the foundation of game AI.

Today—despite the advancements in machine learning and complex physics engines—FSMs remain crucial because games require controllability, not randomness.

This diagram illustrates a finite state machine (FSM) used to simulate enemy AI behavior in a game. The enemy initially enters a wandering state, patrolling the environment until it spots the player. When the player approaches or attacks, the enemy enters an attack state. If the enemy begins to be at a disadvantage, it switches to a fleeing state to escape and ensure survival. When the player leaves, the enemy returns to the wandering state. If the enemy's health is low, it enters a recovery state to restore health, but this recovery process may be interrupted if the player attacks again. After health is restored, the enemy continues to wander, completing a behavioral cycle, thus achieving predictable and plausible game interaction.

Even in today's game industry, where advanced techniques such as behavior trees and machine learning are available, FSMs remain widely used as the "core control structure" for gameplay logic and AI decision-making. They provide a clear bridge between design diagrams and implementation code, support collaboration between designers and programmers, and form the backbone of many hybrid AI systems used in contemporary games.