Periodic Reporting for period 3 - SOFTMANBOT (Advanced RoBOTic Technology for Handling SOFT Materials in MANufacturing Sectors)
Período documentado: 2022-10-01 hasta 2023-03-31
In today’s industry, robots see a widespread implementation for more traditional and repeatable tasks such as assembly and machining while dexterous manipulation of soft and deformable materials is a complex challenge that researchers are still facing.
From surgery to food handling, from packaging to manufacturing, many tasks involving the handling of deformable materials are today performed manually by skilful human operators. The intrinsic complexity of these tasks lies in the high variability of the handled material shape, making it impossible to automate their execution with traditional robotics techniques.
The EU co-funded project SOFTMANBOT started in October 2019. It develops an innovative and holistic robotic system for the handling of flexible and deformable materials within labour-intensive production processes. The developed robotic system will be composed by three main pillars: advanced perception system to track the state of the product and collaborating with human operators; a multi-sensor adaptive control and planning platform used to regulate the product’s state and which will enable human robot collaboration and AI based task generalization; and innovative dexterous grippers with smart mechanical design that will integrate sensors and embody grasping manipulation skills with the soft products.
The SOFTMANBOT findings will contribute to help robots play a bigger role in European factories thereby increasing productivity and job quality.
The SOFTMANBOT project strives to:
• Design a modular and interoperable architecture.
• Develop a robotic perception platform.
• Develop a multi-sensor planning and control system.
• Design and develop intelligent dexterous and low-cost grippers for deformable objects handling.
• Deliver an environmental, economic and social evaluation.
From surgery to food handling, from packaging to manufacturing, many tasks involving the handling of deformable materials are today performed manually by skilful human operators. The intrinsic complexity of these tasks lies in the high variability of the handled material shape, making it impossible to automate their execution with traditional robotics techniques.
The EU co-funded project SOFTMANBOT started in October 2019. It develops an innovative and holistic robotic system for the handling of flexible and deformable materials within labour-intensive production processes. The developed robotic system will be composed by three main pillars: advanced perception system to track the state of the product and collaborating with human operators; a multi-sensor adaptive control and planning platform used to regulate the product’s state and which will enable human robot collaboration and AI based task generalization; and innovative dexterous grippers with smart mechanical design that will integrate sensors and embody grasping manipulation skills with the soft products.
The SOFTMANBOT findings will contribute to help robots play a bigger role in European factories thereby increasing productivity and job quality.
The SOFTMANBOT project strives to:
• Design a modular and interoperable architecture.
• Develop a robotic perception platform.
• Develop a multi-sensor planning and control system.
• Design and develop intelligent dexterous and low-cost grippers for deformable objects handling.
• Deliver an environmental, economic and social evaluation.
The project focuses on the design of robots capable of handling soft and deformable materials in the manufacturing industry, thus reducing the involvement of human activity in these processes. Four industrial partners from four key different manufacturing sectors: toys, textile, footwear and tyres, take part in the SOFTMANBOT project. These industrial partners have their own production lines with an implemented technology which is different from each other. Due to this, the use cases definition and system requirements was accomplished in order to develop the technology which gives a solution adapted to the different cases. Moreover, the performance assessment metrics from the four cases and the global definition of Hardware and Software architecture have been already developed.
One of the pillars of the SOFTMANBOT project is to develop an advanced perception system with the objective of tracking the state of the product and collaborating with human operators. In this field, a wide variety of both visual and tactile sensors have been studied and tested, looking for the solution which fits in the best way the requirements of the different use-cases. In human-robot collaboration the interface of the robot which connects all the sensors and allows the interaction between the human and the machine has been developed.
It is important to control the deformation of the product (shape servoing) and the contact forces applied (contact servoing) in order not to produce any damage in the manipulation of the products. In this field, excellent advances have been carried out. Regarding the shape servoing some geometric models have been studied and in contact servoing an approach based on inverse Finite Element Model (FEM) was developed. A great progress has been also performed in the human-robot collaboration: the State of the Art has been studied and a haptic glove was used for the measurement of the user's motion information, also providing force and vibrotactile feedback to the user. Finally, the State of The Art of task generalisation through robotic skills based on AI has been studied. All of this contributes to the progress of the second pillar of the SOFTMANBOT project.
In the SOFTMANBOT project it is crucial to design robots which can handle soft and deformable products with different shapes. Due to this, the first prototypes of innovative dexterous grippers with smart mechanical design have been designed and developed, considering the specific requirements of the different use-cases. The prototypes of the grippers have been delivered to relevant partners. These partners will give the feedback in order to get the final version of them. Due to the design of the grippers, sensors for tracking the products could be mounted on them.
With the objective of recreating the environment of the four different use-cases and studying the technologies developed in the SOFTMANBOT project, the preliminary recreation of mock-up use cases and the design of the robotic cells have been started. To study the technological developments, simplified tests have been also developed, and the test planning was defined.
SOFTMANBOT is a European project that involves partners from four different European countries and which will contribute to a great development in the industrial sector. Due to this, it is crucial to collect data from the end user in order to minimise the risk of adoption of this technology in industrial companies and also obtaining quantitative indicators that support the potential to bring back production to Europe and the improvement of work environment and safety conditions of the employees. In this field a wide number of questionnaires have been designed and carried out focused on collecting information from both workers and final consumers.
One of the pillars of the SOFTMANBOT project is to develop an advanced perception system with the objective of tracking the state of the product and collaborating with human operators. In this field, a wide variety of both visual and tactile sensors have been studied and tested, looking for the solution which fits in the best way the requirements of the different use-cases. In human-robot collaboration the interface of the robot which connects all the sensors and allows the interaction between the human and the machine has been developed.
It is important to control the deformation of the product (shape servoing) and the contact forces applied (contact servoing) in order not to produce any damage in the manipulation of the products. In this field, excellent advances have been carried out. Regarding the shape servoing some geometric models have been studied and in contact servoing an approach based on inverse Finite Element Model (FEM) was developed. A great progress has been also performed in the human-robot collaboration: the State of the Art has been studied and a haptic glove was used for the measurement of the user's motion information, also providing force and vibrotactile feedback to the user. Finally, the State of The Art of task generalisation through robotic skills based on AI has been studied. All of this contributes to the progress of the second pillar of the SOFTMANBOT project.
In the SOFTMANBOT project it is crucial to design robots which can handle soft and deformable products with different shapes. Due to this, the first prototypes of innovative dexterous grippers with smart mechanical design have been designed and developed, considering the specific requirements of the different use-cases. The prototypes of the grippers have been delivered to relevant partners. These partners will give the feedback in order to get the final version of them. Due to the design of the grippers, sensors for tracking the products could be mounted on them.
With the objective of recreating the environment of the four different use-cases and studying the technologies developed in the SOFTMANBOT project, the preliminary recreation of mock-up use cases and the design of the robotic cells have been started. To study the technological developments, simplified tests have been also developed, and the test planning was defined.
SOFTMANBOT is a European project that involves partners from four different European countries and which will contribute to a great development in the industrial sector. Due to this, it is crucial to collect data from the end user in order to minimise the risk of adoption of this technology in industrial companies and also obtaining quantitative indicators that support the potential to bring back production to Europe and the improvement of work environment and safety conditions of the employees. In this field a wide number of questionnaires have been designed and carried out focused on collecting information from both workers and final consumers.
SOFTMANBOT will produce an outstanding impact to the industrial manufacturing sector. It will help to achieve a higher level of “good quality” production: thanks to the automatization, the process becomes more stable, ensuring a better consistency in the output. Also, it will reduce the frequencies of absences of workers due to health issues, e.g. related to carpal syndrome at the wrist. Moreover, higher level of independence in workers hiring will be achieved, as no need of special qualifications/skills/force is expected, and costs in repairing non-conforming products will be reduced. Finally, the project will produce higher efficiency in demoulding/assembly and better attractiveness for jobs with higher technical environments and less physical tasks.
Main outcomes of SOFTMANBOT will be:
- Complete robotic cell for Textile sector
- Complete robotic cell for Tyre sector
- Complete robotic cell for Footwear sector
- Complete robotic cell for Toy sector
- Reduction of workers stress and increase of job quality when manipulating flexible materials
- Increase of competitiveness in four sectors, which at present, are handwork intensive
- Bringing back to Europe these sectors factories
Main outcomes of SOFTMANBOT will be:
- Complete robotic cell for Textile sector
- Complete robotic cell for Tyre sector
- Complete robotic cell for Footwear sector
- Complete robotic cell for Toy sector
- Reduction of workers stress and increase of job quality when manipulating flexible materials
- Increase of competitiveness in four sectors, which at present, are handwork intensive
- Bringing back to Europe these sectors factories