New research and tools help Internet of Things applications adapt better

In a society built on communication and Internet of Things, computer systems that can adapt to changing circumstances instead of crashing become ever more important. Muhammad Usman Iftikhar's research sheds important light on how one can guarantee that a self-adaptive system is able to achieve its goals.

What if a computer system that controls power distribution could see a problem coming up and reconnect the distribution before a breakdown causes a problem? That's one example of the ideas behind so-called self-adaptive systems – systems that can change their structure or optimize their behaviour to deal with dynamic environmental conditions in which they operate, for instance.

An approach to achieve self-adaptation is using a feedback loop that monitors the situation, reasons about the changes, and adapts the system to meet the goals. An example is an Internet of Things application with sensors monitoring e.g. temperatures. The data packets sent to the user could be lost due to bad weather, wi-fi disturbances etcetera. The feedback loop is placed on top of the IoT application and contains the adaptation logic to increase or decrease the signal strength optimally.

One of the researchers in the field is Muhammad Usman Iftikhar, who recently defended his thesis entitled A Model-Based Approach to Engineer Self-Adaptive Systems with Guarantees. His research is about efficiently providing guarantees in self-adaptive systems, that is, is the system able to achieve its goals?

"I contributed to state of the art with so-called executable formal models of the . These are mathematical models used to rigorously specify and verify software systems behaviour. Another novel contribution is to use runtime simulation and statistical model checking to reason about changes in environment. This supports tradeoffs between available resources and the required accuracy of the results," says Usman Iftikhar.

Self-adaptation can be applied to large scale systems as well as small scale ones, and it's not hard to find examples where it may have a great impact on everyday life.

"For example, we applied self-adaptation in a tele assistance system that provides health support to elderly people in their homes," says Usman Iftikhar.

Usman came from Pakistan to Växjö in 2009, did his second Master in Computer Science here and then continued as a doctoral student. His plan is to continue doing research, focused on using machine learning algorithms for finding the most suitable adaptation options among the thousands that may be available.

"I'm inspired by the challenges you face during research. Sometimes you get results and sometimes you learn. The research never stops. We continue to build solutions for real world problems and that, I think, is the most exciting thing about research," concludes Usman Iftikhar.


Explore further

How neuroscience helps to advance machine learning

Provided by Linnéuniversitetet
Citation: New research and tools help Internet of Things applications adapt better (2017, December 21) retrieved 13 November 2019 from https://phys.org/news/2017-12-tools-internet-applications.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
49 shares

Feedback to editors

User comments