Ambient intelligence: Technologies, applications, and opportunities

Introduction

Computer science is still a new field of science. It has seen many important and rapid changes in its first decade of existence. These transformations have created a fascinating mix of experiences and expectations that are allowing for the creation and deployment of technology to improve our environment's effectiveness. Ambient Intelligence is a new area of research that explores this technical possibility. This section examines the field of Ambient Intelligence. We will be examining the technology that has supported research in Ambient Intelligence (AMI) and how they have been developed. Additionally, we provide an overview of Amy's current applications in practice settings and discuss future opportunities for my research.

Emergence of AmI

In 2001, the European Commission was the first to chart a course for my research. Technology's advancement was a key factor in the birth of AmI. Computers were expensive at first, and they were difficult to use. Computers were a precious and rare resource. One computer could be used by several people. The next evolutionary step was that many users could access the computer simultaneously, so they no longer had to use it in turn. In the 1980s, the PC revolution changed this ratio to one computer per user. With the rise of industry and lower costs, one user was often able to access multiple computers. Today's computing resources are vastly more diverse than those of a few decades ago.

Access to multiple computers doesn't necessarily mean having both a computer and a laptop. The miniaturization and integration of microprocessors have allowed computing power to be embedded in everyday objects like washing machines, microwave ovens, refrigerators, and fridges. They also travel with us (e.g. mobile phones and PDAs) and help us navigate our way (e.g. cars and GPS navigation). 

Computers that can perform more complex computations with less power and adapt the computation to specific tasks are slowly spreading throughout society. Ambient Intelligence was born from this widespread availability of resources.

It is not enough to have the right technology for a science area to thrive. Over the past decades, users' experiences with computers have created an interesting environment where people are more comfortable using them and less afraid of them. This change in society's perception of technology is also driving a shift in how services are delivered. 

This is evident in the decentralization and development of social and health care assistive technologies. The way is now open for AmI systems that can support the care of patients at home and within their communities, as governments and health professionals have begun to move away from a hospital-centric system. All of the components necessary for AmI systems to support the care of patients closer to home, within their communities, are present: the need to spread technology around us, our desire to transform the way we interact with technology, all of the technological knowledge available, and the ability to meet the demand.

Ambient Intelligence isn't a new concept. What is new is the ability to seriously consider it a reality, and a discipline with unique contributions. Many of us have seen science fiction movies in which doors open when someone approaches or computers can identify interlocutors without their names being mentioned explicitly. Although some of these features were impossible with the technology at the time, the industry began to develop features that indicated sensible autonomy for the system.

Many of today's homes are technically smart and can be purchased at a reasonable price. There are many movement sensors and thermostats that control lighting. The bar is now much higher. A society that interacts regularly with such facilities will not be impressed by the ability to link motion sensors to security alarms for detecting intruders.

Recent electronic and computational advances have led to an increase in autonomous semi-intelligent behavior by smart homes. New terms such as Ambient Intelligence were created.

Ambient Intelligence (AmI), as its name implies, is a technology-enriched environment. This includes sensors and devices connected through a network. The system acts as an "electronic butler", sensing features and analyzing the data to determine which actions will be most beneficial for the environment.

These definitions and the features we use to define Ambient Intelligence allow us to see how it compares and contrasts against fields like ubiquitous computing, pervasive computing, artificial intelligence, and artificial biology. AMI systems need to be responsive, adaptive, and sensitive. This highlights the dependence of AmI research on context-aware computing.

Contributing technologies

Amy has a significant relationship with many areas of computer science. The contributing technologies are divided into five categories. The presence of intelligence is a key component in my research. Russell and Norvig defined an intelligent agent. We use this concept. The AmI algorithm can perceive the environment and users using sensors and reason about it using various AI techniques. It then acts on the environment using controllers so that it achieves its goal. We focus on technologies that aid in sensing, reasoning or acting.

Amy, on the other hand, draws from AI but should not be equated with AI. Five key technologies are needed to make AmI a reality, according to the IST Advisory Group. These technologies are not typically covered by AI research, so they will be addressed separately. These include security systems and devices and human-centric computer interfaces. We will then discuss the recent developments in these areas which enhance the development of AmI.

Take action

AMI systems link reasoning to the real world through sensing and then acting. AMI systems can make decisions and influence system users through intelligent and assistive devices. Robots are another mechanism. Science fiction stories have extensively explored the relationships between humans and machines. Turkle points out that robot pets are becoming more common in everyday life, and brings science fiction down to earth. Robotics research has advanced to the point that users no longer have to struggle with getting them to move to a specific location. Instead, they can ask for "bring me some medicine on the counter". These robot assistants can be found in nursing homes, and are a way to provide support for the elderly.

Robots can perform a wider variety of tasks in support of me. They can monitor vital signs and stimulate conversation. Robots can now display more human-like emotions than ever before and even influence human decisions. The museum traffic control project is an example of this. A robot generated cues to allow visitors to go to areas that they would normally avoid. Robots enable AmI systems to be self-mobile and human-like, which allows for greater human interaction and allows AmI systems to have a greater influence on human culture.

Interaction between human and computer

The IST Advisory Group emphasized that AmI must be easy to use to increase society's acceptance. This can be further explained by the need to create human-centric computer interfaces which are context-aware and natural. These are some of the most recent developments in these areas.

The 21st-century ubiquitous computing scenario models are not only dependent on the development and deployment of mobile devices with high capabilities (such as wearable computers or web-phones), but also on the advancement of machine-to-machine computing technologies that allow devices to interact with each other and the network infrastructure. This computing paradigm, also known as sentient or context-aware computing, emphasizes the need for devices to have an intrinsic consciousness about their surroundings and current location.

AMI systems can recognize the context of their environment and can be used to help them. These elements are often more frequent than others in different applications. It is important to know the most recent developments in an environment, such as the time and duration of events, states, and activities. However, identifying the actors is often also important. As we discussed in an earlier section, this may not be an easy task even if you combine more than one sensor (e.g. RFID and movement sensor). You can add a microphone, image processing, and identification infrastructure to your system. For example, you could analyze images taken by a video camera and match them to images of expected occupants. With the subtle qualification, context can refer to many things. Consider, for instance, that we want the system's actions to be taken when "the place" is noisy or "the occupant feels sad". These contexts can be difficult to identify (when is a noisy environment loud enough?). It is also hard to determine with accuracy when someone is sad. While facial recognition, facial gestures recognition, and "body language" recognition all have their place, they are far from being fully functional.

Sources:

https://www.sciencedirect.com/science/article/pii/S157411920900025X

https://dl.acm.org/doi/10.1016/j.pmcj.2009.04.001