Editorial

Sandeep K. Shukla
2015 ACM Transactions on Embedded Computing Systems  
Last week, I attended the International Conference on Embedded Software Systems, which was held in Newark, New Jersey. I found an interesting trend in embedded systems research that is no surprise to anyone, but I thought that it might be worth sharing with the readership of this journal. A glance at the program, including the keynote speeches and panel discussions, confirmed my preconceptions-there is a new trinity in embedded systems research. Most of the activities of embedded systems
more » ... hers, at least those attending the conference, focused on one or more of these three topics. I use the term big data a bit too generically to include machine learning and data mining even when the data is not necessarily "big." The topic of Internet of Things, which already appears to be the subject of many workshops and conferences, seems to have made a sizable impact in the traditional embedded systems community, with some seeing it as the next-generation networked embedded system. Cyber security, of course, has many facets that intersect with embedded systems research-from hardware implementations of cryptographic algorithms to secure cyber physical systems. I have already discussed the views of ACM TECS on the importance of publishing cybersecurity research in the context of embedded systems in the pages of this journal. Machine learning indeed has made very significant strides, in which deep learning seems to surprise us through news articles almost every week-from systems recognizing paintings by master painters to helping brewers of beers improve tastes. However, more interesting to embedded systems research are machine learning techniques to dynamically optimize performance, classifying behaviors of systems between normal and flawed behavior, machine learning based decision procedures to dynamically adapting system parameters based on the load and other factors. Especially when the Internet of Things becomes a reality in improving the lives of people, improving quality of automation systems, and improving transportation system performance, machine learning and data mining will be ready to deliver technologies, algorithms, and possibly products that can be directly used to make those systems perform in the most optimal fashion, adapting to changing situations, and securing the system against hackers who would certainly want to disrupt such systems or try to breach privacy of people who will be connected to such networks. Therefore, I became quite excited by the enthusiasm in this new trinity for embedded systems research. In a panel discussion at the conference with Professors Victor Li and S. Y. Kung, Li talked about a project on monitoring the level of air pollution in Hong Kong, which involved placing sensors at appropriate locations, collecting data, and using data analytics to interpolate and integrate indicators. In my view, this is an impressive project for multiple reasons. First, it is an application of a suitable networked embedded system for a social cause. Second, it is a combination of embedded systems technologies, big data analytics, and perhaps security to avoid any false data injection into such a system to sabotage any policy implications of the findings. ACM TECS invites all researchers working on projects that employ this trinity to consider submitting their unpublished research to be considered for either regular issues of the journal or one of the upcoming special issues focusing on one or some of these three topics. I must emphasize that we are the primary journal for all other
doi:10.1145/2820608 fatcat:jsrdmejawbb47hky7vefgan75m