Ambient intelligence refers to electronic and computing systems which are aware and responsive to actions happening to their environment.
The integration of intelligence in all sorts of embedded systems can lead to the automation of many tasks that are often undertaken by humans, improvement of the quality of life, and more secure environments.
At the heart of ambient intelligent embedded systems usually lies a machine learning/pattern recognition algorithm whose purpose is to make sense and extract meaning from data received from different types of sensors (visual, acoustic, etc.). Using a priori knowledge as well as existing information, intelligent embedded systems can understand and interact with their environment in ways not previously possible and also learn and adapt to dynamic changes in their operating environment. Integration of learning and decision making algorithms into embedded systems can lead to systems that can help doctors diagnose diseases, or even help pilots land airplanes.
Several challenges are associated with the development of intelligent systems related to the development of tools, methodologies, sensor fusion techniques, hardware components, IP cores, and real-time operating systems that will result in efficient and rapid integration of intelligent decision making algorithms into different embedded devices. It is anticipated that through these advancements orders of magnitude of improvement in a wide range of embedded applications, including robotics, healthcare, and surveillance systems.
The challenges associated with all phases of design and development of ambient intelligent embedded systems open up new opportunities for innovation.