Home » Machine Learning/Artificial Intelligence

Artificial Intelligence-based Agent

In this tutorial, we will learn about the artificial intelligence based agent, how it seeks information from its surroundings and how it interacts with it? Also, we will learn the characteristics of an agent and what problems were traditionally faced while designing an agent. By Monika Sharma Last updated : April 12, 2023

What is an Artificial Intelligence-based Agent?

An artificial intelligent based agent is responsible for any work output obtained from the system. It is the main component which while dealing with artificial intelligence because it is the intelligent system which is needed to be developed. An agent can either perceive some information from the environment or can perform some actions on the environment. This work is done through the two main components of an agent: Sensors and Actuators. So, an agent can be anything that can be viewed as:

• Perceiving its environment through sensors
• Acting upon the environment through actuators

Components of Artificial Intelligence-based Agent

1. Sensors

Through which the agent can observe and perceive its surroundings and behavior. All the details that an agent gets from the environment is through these sensors. These sensors can provide information about various things such as about the temperature, weather conditions, traffic, noise, or anything.

2. Actuators

Through which an agent can Act upon the environment and can cause some effect on it. An agent must perform some specific functions or else it is useless. So, any task which a system performs is through its actuators. These can be internal processing chips whose working may be hidden from the user or can be visible moving actuators such as machine arms which may move back and forth.

Example

Let us take an example of a self-driving car.

Here, the agent is our car and its environment is the road it is traveling on and the nearby cars and other things present on the road. So, the car can perceive its environment through traffic proximity sensors, motion sensors, location tracking system and other sensors. The actuators through which the car acts are: the speed system through which it manages the speed of the car, gearbox, steering for handling directions, etc.

Characteristics of an Agent

1. Agent comprises of architecture and Agent program.
2. Agent architecture is the computing device and the Agent program if the function that implements agent mapping.
3. An agent has some abilities, some goals, preferences, and some prior knowledge.
4. An agent has two components: sensors and actuators.
5. From the environment, it perceives past experiences and observations.
6. On the environment, it performs actions.

Traditional Problems of an Artificial Intelligence (AI) Agent

While developing an Artificial Intelligence-based Agent, the vision was to make a computer do things which at present humans can do better. While building an agent, traditionally, the following were the major problems that were faced by the developers:

• Reasoning
• Knowledge representation
• Planning and Learning
• Perception
• Ability to move and manipulate objects

---