Piccolo European Research Project on In-Network Computing
Nine partners from leading companies and universities in the UK and Germany (Arm, Robert Bosch GmbH, BT, Fluentic Networks Ltd., InnoRoute GmbH, Peer Stritzinger GmbH, Sensing Feeling, the Technical University Munich, and the University of Applied Sciences Emden/Leer), kicked off the Piccolo research project on October 15th, aiming to set a shining example of European research collaboration in challenging times.
Piccolo develops new solutions for in-network computing that remove known and emerging deficiencies of edge and fog computing. Piccolo aims to provide new levels of support for innovative applications such as highly scalable vision processing and automotive edge computing.
The research direction in the Piccolo project is about developing in-network computing platforms that are secure and ethical by design, support fine-granular modularisation, are independent of specific network architectures and that provide new levels of performance and robustness by applying a joint optimisation approach for both networking and computing resources.
The Piccolo project is a two-year CELTIC-NEXT project and is funded by BMWi in Germany and Innovate-UK in the UK, as well as the partners themselves.
“From the smallest sensors to more sophisticated applications, the IoT is enriching and growing markets, creating opportunities for real-time insights and autonomous action. The demands on compute and networking will change significantly as we process the data generated from billions of intelligent devices. As part of the Piccolo project, Arm’s research team will look to address these demands by studying the efficiency of serverless technologies, as well as looking at edge compute for IoT performed using in-network computing,” says Chris Adeniyi-Jones, principal research engineer, Arm.
“Today, connected and automated driving are major technological drivers in the automotive domain. This will heavily rely on timely information sharing, generated from in-vehicle and external data sources for decision making. Bosch is tackling the challenges of collecting, processing, storing and disseminating such information efficiently and secure. Piccolo´s vision of bringing resources closer together is a significant step in this direction, and therefore, will have a big impact in the automotive domain,” says Dennis Grewe, Research Engineer at Bosch Research, Robert Bosch GmbH.
“Piccolo will help us to improve the agility and reduce the cost of our market-leading 5G network, improve our internal network management operations, and provide a new way for us to support third party services on our evolving 5G infrastructure,” says Peter Willis, Senior Manager, Software Based Networks, BT Applied Research.
“The next-generation of in-network computing will require secure and decentralized computing infrastructure that can cater to different use cases such as handling the massive influx of sensor data from autonomous cars. We are looking forward to advance the Fluentic Networks platforms for secure distributed computing over heterogeneous networks through the collaboration with the Piccolo partners over the next two years,” says Dr. Yiannis Psarras, Co-Founder and Managing Director at Fluentic Networks Ltd.
“Piccolo will allow us to upgrade our Real-Time Ethernet Switch “TrustNode” to a computing router. Ready for upcoming challenges in public networks, industrial IoT and autonomous driving,” says Andreas Foglar of InnoRoute GmbH.
“We are excited to expand our Erlang based GRiSP platform from IoT devices to the cloud and everywhere in between. For computing to be truly pervasive, it needs to also go In-Network,” says Peer Stritzinger, Founder and Managing Director, Peer Stritzinger GmbH.
“We see a future where low-cost and massively scalable sensing of human behaviours in real-world spaces will be unlocked by the ability to redistribute real-time processing of visual and audio data between edge and server elements in smarter, more dynamic and performance efficient ways. We’re excited to be collaborating with Piccolo partners to explore how network functions could enable this future to deliver trusted and pervasive vision sensing applications,” says Jag Minhas of Sensing Feeling.
“In-network computing has the potential to embrace untapped resources in the environment of mobile users, combining elements of pervasive and cloud computing. This offers on-demand support for user-defined applications to run code, access services, and interact with the surroundings of a user as she moves. Piccolo constitutes another major step in our work on cloudless services for decentralized, privacy-preserving, and secure operation that keeps users in control,” says Prof. Jörg Off of Technische Universität München.
“Finding new ways for computing in the network is a major challenge in 5G networks and beyond, for example for enabling privacy-preserving, decentralized, and scalable next-generation applications. The goals of the Piccolo project are well-aligned with Emden University’s Compute-First-Networking (CFN) research, and we are excited to bring these concepts closer to deployment through our colloboration with the Piccolo partners,” says Prof. Dirk Kutscher of University of Applied Sciences Emden/Leer.