:

The French Health Data Hub, a national data space for health innovation

The Health Data Hub (HDH) is a “public interest group” created by French Law on July 24, 2019 and officially established by order of November 30, 2019, following the Villani report "AI for humanity". The HDH ensures easy and unified, transparent and secure access to health data for public interest research projects, aimed at improving the quality of care and patient support. With its state of the art, highly secure and scalable technological platform, the HDH enables public interest project leaders to mobilise very large data sources, to cross-reference them, and to run complex algorithms. Emmanuel Bacry will present the data space developed by the Health Data Hub and will talk about some of the 50 ambitious projects currently supported by the HDH and contributing to finding tomorrow's solutions to improve all citizens’ health.

Federated Machine Learning in Healthcare and Intralogistics

Training machine learning models requires large amounts of data. Sharing data for larger datasets raises privacy concerns, which has proven to be challenging in practice. A new technique, Federated Learning (FL), can help to overcome this dilemma: training is done locally, and only model parameters are uploaded to a central server.

Cybersecurity in the Internet of Things: Is AI our savior?

Our environment is increasingly being shaped by embedded systems that are connected to each other via the Internet and form the so-called Internet of Things. However, due to the increased connectivity, the risk of cyber-attacks on these systems is huge. Therefore, new solutions are urgently needed to protect these systems from cyber-attacks. This talk will give some insights whether new AI approaches will help to ensure the security of the future Internet of Things.

Protecting the AI: cybersecurity challenges in AI supported applications

The more we rely on an AI's help, the more fatal becomes a manipulation of its outcome, be it by polluting the training data, or by tampering with the classification. Protecting the AI becomes even more important if sensitive data is involved, as is the case in a medical diagnosis. This talk discusses cybersecurity challenges of AI applications, with focus on critical infrastructure like the medical sector.

AI in malware detection

A revolutionary way of malware detection, Combination of intelligent image recognition mechanisms & self-learning neural networks

Mirko Ross

Extortion, data theft, sabotage - when cybercriminals hijack devices on the Internet of Things, the consequences can be devastating. The risk for companies is high, because industrial IoT devices are preferred targets for cyber attacks. Security vulnerabilities due to a lack of updates play a central role in this. To make the Internet of Things more secure, the Stuttgart-based start-up asvin.io has developed a solution that secures and monitors the software lifecycle from development and distribution to operation on devices.

Marius Bleif:

Bitahoy develops a cybersecurity solution for smart home appliances. The Bitahoy Watchdog can be plugged into any home network to monitor and protect all your internet connected devices. The Watchdog contains an intrusion prevention system that can detect and block network attacks in real-time without configuration. The system further provides the user with additional tools to preserve their privacy and security in their digital home, such as filtering malicious or inappropriate content.

Sebastien Garcin:

YZR's SaaS platform, dedicated to non technical business analysts and data workers, helps data teams to automate their data preparation processing. Our proprietary AI and ML based tool is specialised in 3 fundamental tasks : standardisation, labelling and matching. With YZR, business teams easily automate data quality tasks. YZR makes data projects possible : smart analytics, sales forecasting, better consumer experience, and more.

Dr. -Ing. Duc Tam Nguyen

The AI startup from Karlsruhe "Aimino Tech GmbH" improves data quality to create more robust and powerful AI systems. We provide data tools to generate new images and videos for Computer Vision Use-Cases. Consequently, data scientists can maximize the economic value of their AI systems in their business cases.

The Challenges of AI at the Edge – Towards a Systematic Discipline for AI Systems Engineering

Many AI systems that interact with “physical reality” fail or get stuck after initial prototypes and before large-scale deployments. Changing this situation is the goal of the Competence Center KI-Engineering (CC-KING). The talk uses actual industrial AI applications to showcase the challenges commonly encountered for AI at the Edge. The talk concludes with current results to overcome these challenges and to establish the use of AI in complex systems as a systematic and dependable engineering practice.

Edge AI, a new battleground for AI and opportunity for European manufacturers

In this presentation, we will examine how Edge AI technology can help European manufacturers leverage the value brought by AI technology without necessarily depending on cloud-based AI technologies and therefore without sharing their data with there.

Anomaly Detection and Natural Language Processing in Wastewater Management

Small and medium-sized enterprises must continue to drive digitalisation in order to remain competitive. The Industrial Internet of Things (IIoT) enables companies to gain live insights into their equipment processes and to detect impending failures or anomalies at an early stage. Furthermore, information retrieval and natural language processing (NLP) approaches support field technicians retrieving relevant documents and initiating optimal repairs. In collaboration with KROHNE, the leading manufacturer of process measurement technology, and the Service-Meister research project, we develop a solution that links water management to the Internet of Things (IoT) and also enables field technicians retrieving most relevant documents and initiating optimal services and repairs.

Life-ready AI is here

At GML we are in the business of life-ready AI. Artificial Intelligence that feels far from artificial. Brain inspired chips that respond like humans do. GML is proud to be unveiling a future where artificial meets reality with its GrAI VIP (Vision Inference Processor). A ground breaking chip that acts and reacts in real-time, ready for life.