'Critical systems will always prioritize reliability over performance'

We were delighted to chat about safety and reliability of critical computer systems with Dr. Lalit Kumar Singh (Nuclear Power Corporation of India Limited, Bhabha Atomic Research Centre), one of the keynote speakers at ICFITT 2017, 1st EAI International Conference on Future Internet Technologies and Trends, which will take place between August 31 – September 2, 2017 in Surat, India.

Could you summarize the scope of your current work and what you are coming to share with everyone at ICFITT 2017?

I deal with the design and development of computer-based safety of critical systems. Apart from developing functional requirements, I address quality attributes for these software systems by performing reliability, safety, and security analysis. Sensors and actuators work on heterogeneous and distributed networks. Due to safety significance of such systems, these need to be designed carefully to ensure their safety requirements. However, digital systems have some unique characteristics, such as used software, and may have different failure causes and/or modes than the analog systems; hence, their incorporation into life critical systems, safety analysis entails special challenges. The strategy to perform safety analysis using Petri net for foolproof design will be discussed with a case study.

What do you see as the biggest challenges that future internet technologies are currently facing?

I feel reliable and safe communication is the biggest challenge that future internet technologies are facing. To meet this requirement, we need to ensure software safety and software reliability at least in an analytical manner, if not quantitatively. If you are dealing with the safety of critical systems, you give the highest priority to the reliable and safe transactions rather than performance (response time).

How do you see research and industry moving forward and tackling these challenges?

Internationally, several researchers, academicians, and professionals are continuously working to devise robust methods to ensure software reliability and safety. There are more than 250 models available for software reliability but unfortunately, no method is generic enough so that it can be used in all the types of software. The behavior or failure rate of the different types of software is different due to their failures in different ways. The related community is moving forward to tackle such challenges and looking to devise a standard approach.

What would you say are the main trends in this area that are showing promise?

Software reliability growth models, models based on Bayesian approach, mathematical formalism like Markov chain and Petri nets, Artificial Neural Network models are the main trends in this area that are showing promise.

ICFITT 2017 is still accepting papers! Find out more here.