People of ACM - Valérie Issarny

November 17, 2020

Why is this an exciting time to be working in middleware, with a focus on mobile applications?

Mobile applications impact such a diversity of domains and users, spanning the professional and the private spheres. Researching the supporting middleware then comes with a multitude of fascinating challenges. The main one is addressing the core function of the middleware, that is, enabling interoperability. Indeed, the middleware must overcome the heterogeneity associated with the variety of underlying software, hardware and network. And, although standards keep emerging, there will always be a technology coming up to disrupt the standard.

As for the research we are currently doing, what is the most exciting for us is the blending of the virtual with the physical world that the Internet of Things advocates. And mobile computing has a key role to play, especially with the increasingly rich set of sensors that mobile platforms are embedding. The research question that arises is: How do we benefit from the mobile IoT to inform our knowledge of the physical world in a way that guarantees privacy and the relevance of the contributed data? Then, instantiating the research question at the middleware level helps to provide solutions that are generic enough so that they benefit various applications and may be deployed over a large diversity of platforms. For all these reasons, it is definitely an exciting time for middleware research focused on mobile applications, and it will certainly be so for some time.

Will you briefly tell us about the goals of the FP7 ICT FET IP CONNECT project, for which you were recognized by the French Ministry of Research?

The perspective of the CONNECT project, which the CONNECT team advocated (and I am still advocating), is that although a standard is a relevant piece of solution to overcome the heterogeneity that is intrinsic to any distributed software system, it will never be the solution. New technologies and standards will always keep emerging. Furthermore, distributed software systems increasingly aggregate highly heterogeneous components, and a single standard cannot cover them all efficiently.

So, the CONNECT project took for granted that there does not (or may not always) exist a built-in solution to reconcile the heterogeneity of networked systems that need to interact. Instead, the long-term vision of the project was to address the interoperability requirement through “emergent middleware,” that is, the on-the-fly synthesis of the middleware in the form of software mediators required for the networked systems to be able to interact. We then undertook supporting research on the underlying theory, and on on-the-fly learning, synthesis and deployment of software mediators. While we were able to elicit the foundations and implement an early prototype of emergent middleware enablers, the overall vision is still on my research agenda. For instance, we have been recently focusing on enabling emergent middleware for IoT-based systems that involve highly heterogeneous components.

Recently, you have been working on developing “participatory systems” using new middleware solutions. Will you give us an example of one of these projects, and how participatory systems might be part of the fabric of smart cities in the near future?

Participatory systems aim at easing the civic engagement of citizens based on the increasing connectivity of people. It is not about denying the issue of a digital divide, which requires political actions. It is about acknowledging that people are increasingly digitally connected and may be willing to get involved in, for instance, public policymaking through digital media. Obviously, social media have been playing a key role here. However, it is important for people to be able to engage with whatever technology they use and like, and not be dependent on using some proprietary solution. Here again, middleware that provides interoperability is a valuable tool. In particular, we are working on the service bus paradigm that allows people to interact across distinct social media.

Another example of participatory systems we have been working on is supporting urban crowdsensing for environmental pollution monitoring. In general, participatory systems enable citizens to engage in the fabric of their cities by contributing knowledge, ideas and even projects for a better urban environment. This includes providing software platforms that support participatory budgeting campaigns, which are getting increasingly popular around the world. We investigated these kinds of platforms as part of a collaboration with the CITRIS Institute at UC Berkeley. We learned that developing participatory systems comes with significant challenges, especially for middleware research. One challenge is to offer the software infrastructure that fosters the engagement of people for the social good at the internet scale, and thus deals with issues of scalability, quality of experience, privacy, security and trustworthiness.

How would you like to see the ACM Europe Council grow in the coming years?

ACM is mainly known in Europe for its conferences and publications. At the same time, European researchers and practitioners know less about the full range of ACM. So, one ambition we have is for ACM to be considered more broadly, and especially by European educators, researchers, and professionals. I believe we need to do a better job of communicating about ACM activities throughout Europe. This coincides with having European contributions to computing acknowledged, starting with increasing the participation of Europeans in various ACM initiatives, including ACM prizes and awards.

As EIC, what will be some of your key goals for ACM TAAS?

With “the world becoming digital,” most computing systems need to embed some kind of “autonomous and adaptive” capability. As such, TAAS is multidisciplinary in nature and concerns various communities. One of my key goals is that the researchers developing solutions to the autonomy and adaptation of systems seek to publish their research in TAAS, instead of a journal of their “home community.” This is a challenging ambition, especially with a journal that is not yet established as a leading journal. Toward that end, I would like to foster the publication of special issues and surveys that highlight the significance of “autonomy and adaptation” for present and future systems. In any case, I have no doubt that the Editorial Board will succeed in making TAAS a leading journal!

With the pandemic, a good deal of our activities have migrated online, and this may possibly be conveniently leveraged, even for journal editions, by considering video presentation/introduction for some papers. In general, we also hope to leverage social media to enable more interaction with contributors. Also, whenever relevant, the publication of related software and datasets should be fostered and supported. I do believe the journal may benefit from a large online presence.