Advanced Research Initiatives: Focus on the Future
Advanced Research Initiatives: Focus on the Future
CA Technologies is launching three new projects to enable smart IoT systems, including the ALOHA, ENACT, and PDP4E systems.
Join the DZone community and get the full member experience.Join For Free
Digi-Key Electronics’ Internet of Things (IoT) Resource Center Inspires the Future: Read More
CA Technologies announced that its Strategic Research team will collaborate on three projects funded by the European Union (EU) Horizon 2020 program. These projects will enable secure and smart IoT systems that will underpin a more trusted and connected world, exploring both the development of algorithms and tools required to build compliance into software development processes.
Dr. Maria Vélez-Rojas, research scientist; Dr. Victor Muntés-Mulero, V.P. of Strategic Research; and Dr. Steven Greenspan, research scientist, at CA technologies further explained the different initiatives they are working on and the issues they raise around privacy, security, bias, and the collaboration between robot and humans — cobots.
CA Strategic Research scientists explore new technologies, applications, and platforms, like IoT, robotics, Artificial Intelligence (AI), and more through diverse R and D efforts in partnership with leading research communities in academia, government, and beyond.
For IoT, the possibilities are limitless —from smart cities to transit and infrastructure. Furthermore, IoT systems are highly complex, layered, and fundamental scientific challenges centered around system resiliency, trust, and continuous innovation.
“To realize the massive promise of an IoT-driven world, we must solve complex challenges,” said Otto Berkes, Chief Technology Officer, CA Technologies. “These hurdles must be overcome before we can deliver IoT systems that can provide valuable and trusted data, be adaptable and open to new technologies — systems that haven’t even been invented yet.”
ALOHA for Deep, Secure Learning
A smart security system that gathers data from IoT devices needs to ensure that the AI-based decisions from that data are not producing biased results. The Adaptive Learning on Heterogeneous Architectures (ALOHA) project aims to improve human decision-making in IoT applications.
ALOHA will explore how a type of AI, known as "deep learning" or DL, can be embedded in IoT applications to imitate biological neural networks and acquire human-like learning capabilities. This work also exploits the increasing compute capacity of edge devices that provide an entry point into enterprise or service provider core networks, like routers and servers, and their ability to support the execution of AI algorithms.
CA Technologies’ security expertise will contribute to a new depth of understanding to this project, which is critical to creating bias-free AI. “CA’s goal through ALOHA is to learn from experience and react autonomously to a surrounding environment while avoiding AI bias. Human rights and fairness are one of the main building blocks of future AI systems. It is our responsibility as humans to drive technology with human rights and fairness in mind," according to Muntés-Mulero.
ENACT for Smarter IoT
Today, smart railway systems require an IoT system comprised of multiple applications. Exposing vulnerabilities early on in the software development process could prevent a major system failure. The ENACT project aims to drive faster innovation across trustworthy, smart IoT systems through new development, operation, and security applications that span across IoT, edge, and cloud infrastructures.
In addition to developing smart home e-health applications, this pioneering research will also help to build AI-based self-diagnosis for smart operation, which will eventually enable trains to “learn” and predict anomalies. These “smart trains” will also need advanced simulation models to ensure IoT system and application security as edge devices are continuously added in the future.
As a leading partner in the project, CA Technologies will push the boundaries of continuous delivery, agile operation, and security technologies and processes designed for smart IoT systems.
“In the future, all devices will ultimately be connected. ENACT research will help transform the understanding of how trusted, smart IoT systems can be developed and operated in our fast-changing environment,” says Muntés-Mulero. “CA’s deep expertise in the enterprise, combined with insights from our Strategic Research, aims to contribute to a more trusted, connected world and transform concepts into breakthroughs.”
PDP4E for Better Privacy and Personal Data Protection
GDPR is a new and comprehensive law that requires businesses and governments to protect the personal data and privacy of individuals in the EU. (This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No 787034.) The regulation has a global impact and cuts across all sectors with non-compliance potentially resulting in massive fines. GDPR is expected to set a new standard not only for consumer rights regarding their data, but also companies challenged with putting systems and processes in place to comply.
The Privacy and Data Protection for Engineers (PDP4E) project will address the protection of personal data and explore how to better enable the development of code that allows companies to comply with GDPR, by enabling engineers with tools to integrate data privacy during the development process (or “Privacy by Design”).
Research for PDP4E will support the development of tools and processes that move data privacy left, facilitating GDPR-compliance for organizations as they create new applications. The research will also explore algorithms for classifying documents with personal information and to simplify the detection of the content that is subject to data protection regulation.
"We must share data for our economy to work, but we must do so while preventing inferences that violate GDPR. As such, we are creating a framework and tools for privacy by design," according to Greenspan.
In collaboration with the EU, CA’s research creates opportunities that accelerate ideas to outcomes. Projects span next-generation infrastructure, security and privacy, and new business models and processes. Muntés-Mulero continues, “For example with ALOHA, CA will also focus on exploring how agile methodologies can be used when embedding deep learning in our applications.” CA’s projects are conducted through the Office of the CTO and led by a team of distinguished researchers, engineers, and staff.
Cobots — the Droids You're Looking For
A fourth problem the team is working on, presented by Dr. Velez-Rojas, involves the collaboration between robots and humans — cobots. Humans are creative problem-solvers, but they often face physical and safety limits that stop them from reaching their full potential. Cobots are designed to support humans in a seamless and safe process, maximizing productivity in complex and continuously changing environments.
Cobots are necessary where errors matter. Sensor information and pattern recognition provide cobots with contextual awareness. By fusing all sensor signal, cobots classify objects and situations. However, inaccurate classification can result in harm to humans.
The initial research associated with this project revealed improved accuracy in cobot activities
How do we communicate what we care about? Human-to-cobot communication requires simple, high-level commands rather than explicit and customized programming. Human intervention requires context acquisition, and, in many cases, humans are required to act quickly. Cobots need to understand what humans care about and decide how to respond.
Cobots also improve the trustworthiness in IoT environments. IoT creates an expanded attack surface, however, it's not just about security. Sensor miscalibration, faulty installations, low batteries, and environmental disruptions can cause problems. Data error can have large consequences (i.e. Air France Flight 447 crash on June 1, 2009, where all three external airspeed sensors gave different readings and human intervention was not able to overcome the errors). We will continue to learn and prevent situations like this going forward.
Opinions expressed by DZone contributors are their own.