Workshop - A&T Next Speech Day - part 3

Start-up presentation:
Sqrrl - Rescomp Engineering Srl - BRD Innova - Vibes srl - Onirys Srl - Anypath S.r.l. - Videam - Niuexta Srl -Magys Drive srl

Software defined automation on M8
The next generation of computing systems for automation must meet new requirements: workload adaptability, interconnection with other systems, cybersecurity and functional safety. M8 is the operating system for critical and software-defined computing platforms. It combines virtualisation and containerisation, a real-time microkernel with the highest levels of cybersecurity and functional safety, compatibility and Linux experience. The result is a software platform for virtual PLCs and flexible, efficient, deterministic and secure edge AI.
Marco Solieri, SQRRL - CEO & Founder Sqrrl

Towards the industry of the future: innovative strategies for monitoring lightweight structures made of advanced materials
High specific performance, the ability to design the material according to the application, and great design freedom make advanced composite materials a key technology for the development of lightweight, efficient, high-performance structures. Their use is today continuously growing in the automation, automotive and aerospace, and energy sectors. However, alongside the growing ability to optimise performance and weight, a new industrial need has emerged: knowing the structural integrity status during operation, early detection of any malfunctions and monitoring their evolution over time.
The presentation will introduce some advanced methodologies for Structural Health Monitoring (SHM) of composite materials, polymers and lightweight structures, showing how the integration of measurement techniques, physics-based models, data analysis and artificial intelligence makes it possible to transform acquired data into information about the actual operating condition of the structure.
Particular attention will be paid to an innovative approach for composite materials: using the structure itself as a sensor. By exploiting the electrical properties of the material, it is possible to detect and localise damage phenomena and progressive performance degradation, reducing or eliminating the need to introduce additional sensors. The component is therefore no longer just a structure to be monitored, but becomes an active part of the monitoring system. A similar approach can be implemented in the case of doped polymers.
The objective is to transform monitoring into a concrete tool for designing and managing the component throughout its entire life cycle. Knowing the actual state of the structure and its evolution during operation makes it possible to reduce over-design margins, designing lighter and more efficient components without compromising their safety and reliability. At the same time, the availability of information on the actual condition of the component makes it possible to go beyond conventional maintenance logic, moving towards condition-based maintenance strategies, in which inspections and interventions are planned on the basis of the real conditions of the structure.
Prof. Michele Zappalorto, Managing Director Rescomp Engineering Srl

Kraken RT30, the rotary palletiser designed to automate the loading and unloading of items.
Bordignon Stefano, owner and administrator BRD Innova

Atacod

Predictive maintenance and other real-world AI use cases for manufacturing
Real-world cases of predictive maintenance through vibration analysis: failure and maintenance forecasting in industrial machines, as well as AI applications in manufacturing: anomaly forecasting, critical operating conditions, accident prediction with computer vision for workplace safety.
Matteo Pozzi, administrator, Vibes srl

Onirys Srl
Speaker: Emiliano Mari, Onirys S.r.l. (brand Agent24), Chief Executive Officer

Physical AI: Flexible Robotics for High-Mix, Low-Volume Production
Manufacturing has been automated for decades, with millions of robots already deployed in production facilities. However, the rapid shift towards high-mix, low-volume production is making traditional automation increasingly rigid, complex and costly to adapt. Programming a robot still takes days or weeks, and every brand uses proprietary software and programming languages. Added to this is a further limitation: complex processes such as welding, coating and finishing require years of experience and process expertise, today held by only a few specialised engineers and difficult to transform into structured, shareable and reusable company know-how. Anypath is the software layer that enables flexible production at scale. Our foundation model learns how experts perform industrial processes and uses this know-how to make them adaptable in seconds. The result is a new generation of industrial automation, capable of rapidly adapting to production variability and generating savings of over hundreds of thousands of euros per year for each robot.
Mirco Bucci, CEO - Anypath

Quality control with AI: learning the conforming product to recognise anomalies
How can quality control be automated when defects can take different forms and are not all predictable?
VIDEAM presents an AI-based software library that learns the characteristics of a conforming product in just a few steps and identifies in images what deviates from the expected characteristics.
The presentation will illustrate the anomaly detection approach applied to industrial visual inspection and the role of hardware in making defects recognisable. Alongside its software expertise, VIDEAM also handles the design of the vision system and the selection of cameras, optics and lighting suited to the application.
The objective is to integrate these technologies into production processes, delivering turnkey solutions: from image acquisition to anomaly detection, through to interfacing with the line.
Tommaso Scuccato, Co-founder and CTO, VIDEAM SRL

The production line is automated, the office is not: AI removes the bottleneck for SMEs
Robots have made the production department more efficient. Office activities are often not optimised: quotes built manually by reading drawings and specifications, technical offices searching for answers in manuals, documentation to be updated in line with continuously evolving laws and regulations. In ten minutes, NIUEXA, an innovative company developing artificial intelligence agents for manufacturing, shows with concrete examples how AI reads technical documents, connects to existing management systems and prepares the work, leaving the final decision to people.
Gregor Maric – CEO & Co-Founder, NIUEXA Srl

Magys Drive: the contactless gearbox
Magys Drive redefines motion control for Industry 5.0 with the first contactless magnetic (gear)reducer equipped with an integrated torque sensor. By replacing traditional mechanics with magnetic interaction, our patented technology eliminates friction, wear and backlash, ensuring maintenance-free and lubricant-free operation. Born from the excellence of the Scuola Superiore Sant'Anna, Magys Drive unlocks higher levels of precision, quietness and sustainability for a new generation of cobots and robots for industrial automation.
Eng. Andrea Pergolini, CTO, Magys Drive srl

Hosted by: Giovanni Turri