Con esta herramienta te facilitamos un acceso a todas las ofertas y demandas de tecnología europeas y a búsquedas de socios para participar en propuestas europeas de I+D publicadas en la red Enterprise Europe Network, pudiendo filtrar los resultados para facilitar las búsquedas más acordes con tus necesidades.

¿Quieres recibir estos listados de oportunidades de colaboración en tu correo de forma periódica y personalizada? Date de alta en nuestro Boletín

Para optimizar los resultados de la búsqueda, se recomienda utilizar términos en inglés.

Deep learning-based poka-yoke process monitor

Resumen

Tipo:
Oferta Tecnológica
Referencia:
TOIT20200921001
Publicado:
07/10/2020
Caducidad:
08/10/2021
Resumen:
An Italian SME has developed an innovative system which makes any rule-based process error-free. Designed for real-time checking of human-machine interaction, human or machine actions within manufacturing. It guides human and/or machine actions with human-like collaborative vision in order to improve efficiency, production quality and safety. The firm is open to technical or commercial deals with technical assistance with end users coming from the industry and manual processes to be managed.

Details

Tittle:
Deep learning-based poka-yoke process monitor
Summary:
An Italian SME has developed an innovative system which makes any rule-based process error-free. Designed for real-time checking of human-machine interaction, human or machine actions within manufacturing. It guides human and/or machine actions with human-like collaborative vision in order to improve efficiency, production quality and safety. The firm is open to technical or commercial deals with technical assistance with end users coming from the industry and manual processes to be managed.
Description:
An Italian SME has developed and patented an innovative system which makes any rule-based process error-free.
The system has been designed for real-time checking of human-machine interaction or human and machine actions within manufacturing. The system guides human and/or machine actions with human-like collaborative vision in order to improve efficiency, production quality and safety.
The firm is open to technical or commercial deals with technical assistance with end users coming from the industry having manual processes to be managed.
Errors and anomalies are unavoidable in any manufacturing process. But if mistakes are caught at the right time, the consequences can be prevented. The initial removal of defects reduces errors costs, in many cases down to zero.
This patent-pending solution uses 3D machine vision and artificial intelligence, applied in collaborative mode, to understand the events that occur within a certain workspace.
Thus, the system puts a motion-step method-based poka-yoke in place, checking if all the tasks of a given process have been performed correctly, in the preset order and times.
The solution is a real-time process monitoring system that helps users to achieve zero-defect manufacturing and maximize efficiency, quality and safety.
Optimizing time-and-motion means improving efficiency and productivity, enabling the staff to manage and execute production processes correctly. This is the way to get great savings, eliminating all waste and minimizing costs. But the optimized work cycle is often difficult to respect, for humans and sometimes also for machines. The system works alongside those who work, whether this is the man or the machine or both at the same time. It collaborates with them, in a non-intrusive but inflexible way: it monitors and guides their work in order to maximize efficiency, quality, and productivity while preserving the safety of operators as well as the performance and integrity of the machines since it can intervene promptly in case of danger.
The proposed solution takes furthermore advantage of an exclusive combination of cutting-edge technologies, making it unique in terms of performance and flexibility of use.
Deep learning makes the system capable of recognizing multiple elements moving in the scene - including objects which can vary in shape - without getting confused by distortions, different perspectives, variable contrasts or objects which look not really sharp. Furthermore, the learn-by-example algorithms that characterize deep learning allow the AI quotient to increase indefinitely, with no need for human intervention.
3D machine vision grants the system the ability to see with maximum realism what happens at any moment, with the spatial perception typical of human vision. So, it is possible to obtain:
- A real-time contactless control, free from changes in performance over time
- A constant, impartial, evaluation capability that allows to classify both the quality and efficiency of your process.
Setting up a process control with this system is easy and intuitive, in fact, a wizard suggests how to set up the system. It is not necessary to write even a line of programming code: the approach is totally visual and does not require specific competencies. After the configuration phase, the system turns the attention to the workspace and interacts with human operators, machines or both, providing directions, reporting any mistake in the performance of the process under control and nipping potentially dangerous situations in the bud. Thanks to the ability to analyze seamlessly, even multiple simultaneous events, this solution offers unprecedented reactivity in the event of anomalies.
Advantages and Innovations:
For quality purposes:
Industrial manufacturers must always solve two problems and bear two costs: identifying defective parts and removing them from production lines. Wouldn´t it be better to only get compliant items and be free from inspecting the whole production and to discard bad parts? This system picks out the actions that cause defects, stopping them in their tracks thus preventing the creation of non-compliant products and saving a lot of time and money.

For safety purposes:
The ever-closer cooperation between man and machine, the centerpiece of the breakthrough Industry 5.0, harmonizing human and artificial intelligence aims to create new forms of added value in production but poses new challenges on the front of the safeguard in the workplace. The system monitors in real-time the safety of human operators as well as machines or both and acts as a supervisor of their activities, establishing a climate of optimum health and safety with all the benefits that result in terms of flexibility, speed and competitiveness.

For predictive maintenance:
By opting for a model based on predictive maintenance on the factory floor is a strategic choice that has a positive impact on business continuity, but it is not always possible to create a direct connection with machines under control in order to acquire the necessary data. This solution replaces material connections with observation, characterized by continuous-time and three-dimensional vision, through which it detects drifts in motion, position and pace of work. The seamless analysis of such data allows the system to predict and thus prevent failures that could generate damage, machine downtime and safety issues.
Stage of Development:
Already on the market
IPs:
Patent(s) applied for but not yet granted

Partner sought

Type and Role of Partner Sought:
The company is interested in commercial agreements with technical assistance with end users coming from the manufacturing industry having manual processes to be managed.
The company is also interested in technical cooperation agreements with technical partners willing to propose integrated systems to their final clients.

Client

Type and Size of Client:
Industry SME <= 10
Already Engaged in Trans-National Cooperation:
No
Languages Spoken:
English
Italian

Keywords

Technology Keywords:
01003003 Artificial Intelligence (AI)
01001001 Automation, Robotics Control Systems
10001002 Evaluación de riesgo e impacto medioambiental
02003005 Information processing & Systems, Workflow
02003001 Process automation