Automated and emerging technologies · ⁨자동화 및 신기술⁩

An automated system 自动化系统 is a mix of software and hardware that senses and responds to data in its environment 环境, with no need for human intervention 人工干预. In other words, it works on its own.

Examples of automated systems:

• a central heating 中央供暖 system;
• a chemical process 化学过程 in a factory;
• a greenhouse 温室 (for growing plants);
• a car park barrier 停车场道闸.

Sensors, microprocessors and actuators

Three parts work together in an automated system.

The cycle works like this:

1. The sensor sends data to the microprocessor.
2. The microprocessor compares this data with stored values 存储值 and makes a decision.
3. The microprocessor sends signals to the actuators to take action.

For example, in a greenhouse: a temperature sensor reads the heat; the microprocessor compares it with the wanted value; if it is too hot, the microprocessor signals an actuator to open a window.

The syllabus names many sensor types 传感器类型, each reading one physical quantity – for example an accelerometer 加速度计 (movement or tilt), a humidity 湿度 sensor (water in the air), and a proximity 接近 sensor (a nearby object):

To pick a sensor for a scenario, choose the one that reads the right physical quantity: a fish tank needs pH and temperature sensors; an automatic door needs an infra-red / proximity sensor.

Advantages and disadvantages

Automated systems are used in many situations, such as industry, transport, agriculture 农业 (farming), weather (gathering data), gaming and lighting.

Worked example. Describe how an automatic greenhouse holds the temperature at 25 °C. A temperature sensor continuously measures the temperature and sends its reading, converted to digital by an ADC, to the microprocessor. The microprocessor compares the reading with the stored value of 25 °C. If the temperature is above it, the microprocessor signals an actuator to open a window or switch on a fan; if it is below, it switches on a heater. The whole loop then repeats continuously. Two marks nearly always sit on the words "compares with a stored (pre-set) value" and "the process repeats" - a description that stops at "the sensor tells the computer" leaves them behind.

Robotics 机器人技术 is the branch of technology that deals with the design, building and operation of robots 机器人.

A robot has these characteristics:

• a physical structure 物理结构 (a mechanical 机械的 body, such as arms);
• electrical components 电子元件 (motors, sensors and wiring);
• programmable instructions 可编程指令 — it follows a program that can be changed.

What robots do

Robots can perform many roles, for example:

• factory equipment 工厂设备 — building cars, welding, moving heavy parts;
• domestic appliances 家用电器 — such as a robotic vacuum cleaner 吸尘器.

A key limit of a robot is that it cannot think for itself. It cannot make its own choices when something unexpected happens.

Artificial intelligence 人工智能 (AI) is the simulation 模拟 of human intelligence by computer systems. This means a computer doing tasks that normally need human thinking.

Characteristics of AI

AI systems usually have these features:

• they collect data and the rules 规则 for using that data;
• they have the ability to reason 推理 (work things out using the rules);
• they have the ability to draw conclusions 得出结论, which may be approximate 近似的 (a best guess) or definite 确定的 (certain);
• they have the ability to learn 学习 from data;
• they have the ability to adapt 适应 — to change their behaviour as they get new data.

Examples of AI

• expert systems 专家系统 — software that gives advice like a human expert (for example, helping a doctor with a diagnosis);
• natural language processing 自然语言处理 — understanding human speech or text;
• self-driving cars 自动驾驶汽车.

Inside an expert system

An expert system stores human expertise and reasons from it. It has four parts:

• a knowledge base 知识库 – the stored facts about the subject;
• a rule base 规则库 – the "if… then…" rules a human expert would apply;
• an inference engine 推理引擎 – applies the rules to the facts to reach a conclusion;
• a user interface 用户界面 – asks the user questions and shows the result.

The interface collects facts from the user; the inference engine runs the rule base against the knowledge base; and the system outputs a diagnosis or a probability – as used to help a doctor diagnose an illness or to find a car fault.

Machine learning

Machine learning 机器学习 is a program that improves its own performance from experience, without being explicitly reprogrammed. It finds patterns in data and adapts: a search engine gets better at ranking results, a voice assistant recognises spoken commands more accurately, and a robot vacuum gradually learns the layout of a room.

Narrow, general and strong AI

• Learn the control loop: a sensor measures a quantity → the microprocessor compares it with a stored value → an actuator takes action; then the loop repeats.
• Do not confuse the parts: a sensor sends data in; an actuator causes movement out.
• A robot follows programmable instructions and cannot make its own decisions when something unexpected happens.
• AI simulates human thinking: it collects data and rules, reasons, draws conclusions, and can learn and adapt.
• Know the AI levels: narrow (one task; all AI today), general (any human task), strong (self-aware; does not exist yet).