IkusmenA - Mobile robot with artificial vision navigation for industry
March 2021 - December 2021
Project subsidized by: Dpto. Desarrollo Económico, Sostenibilidad y Medio Ambiente of Gobierno Vasco and European Union (FEDER)
AbstractThe main objective of the IkusmenA project is to develop the technology of an autonomous mobile robot with integrated artificial vision and manipulation capacity for use in assembly and precision assembly tasks, such as those carried out on final assembly lines in the automotive industry. and manufacturing.
This technology implies that the mobile robot itself scans the work area, locates its position, plans the trajectory and is guided through its navigation system, safely avoiding obstacles, whether fixed or mobile, which it detects in its path by means of cameras. vision, with the ability to adapt to dynamic changes that may occur in the work environment. Likewise, artificial vision, enabled by image processing through convolutional neural networks, allows the precise identification of the work environment, which enables the robot through its manipulator arm to execute precision tasks in assembly lines, both static and dynamic.
Autonomous mobile robot technology capable of navigation and precision manipulation by artificial vision is currently at low levels of technological maturity and is not deployed in the industry except for some pilot experiences based on prototypes. Therefore, a robust and reliable mobile robot with navigation and manipulation capabilities based on artificial vision represents a novel technological solution that meets the needs of the demanding manufacturing industry oriented to the "Lean manufacturing" philosophy. This development can be considered a New Product since it is a disruptive technical solution that enables the automation of the final assembly phase of production chains, the only phase of product manufacturing that is currently still carried out by manual labor in most of the automotive industries and manufacturing industries in general.
InnovationThe new product to be developed in this project is aimed at developing an automation solution that allows overcoming the current technology of the final assembly lines and its limitations and barriers to automation. The main contributions of this new product in terms of functionality and quality with respect to the current state of the art in final assembly tasks in the industry, are the following:
- Improved reliability of assembly processes: Tasks and processes are executed reliably by mobile robots, relieving personnel of the most tedious, monotonous and effortful tasks, and allowing human resources to contribute ideas and knowledge in other more value-added tasks. The automation of the final assembly lines implies precision and high repeatability, ensuring that the product is manufactured with the same quality specifications every time.
- Quality improvement: Manual or automated assembly lines can perform different tasks to achieve specific manufacturing specifications. However, achieving the same level of quality consistently requires performing tasks and processes with specific repeatability. Autonomous mobile robots guarantee methods of high repeatability, which allows guaranteeing the quality standards defined in the final product assembly lines.
- Cost savings: For most manufacturers, assembly line personnel costs represent one of the highest direct costs. Automating assembly lines using autonomous mobile robots enables tight process control and a predictable and maintained level of quality, all while reducing direct costs. With the robotization of lines, the assembly line staff can be relocated to tasks that provide much more added value, which in turn allows improving the competitiveness of the industry.
- Increased traceability: The automation of the final assembly allows the perfect traceability of each unit of manufactured product. Upon detection of an error or defect, it is possible to review all the previous processes to precisely identify the origin of the error and implement preventive measures.
- Productivity improvement: The robotization of the final assembly lines allows for a notable increase in productivity, by increasing manufacturing speed and reducing costs, which directly affects the profitability of the manufacturing industry and its competitiveness in the market
- Improved operational flexibility: Assembly line automation with autonomous mobile robots enables easy adaptations to different work environments, changes in those environments, and changing tasks. All of this eliminating the long learning curves, and ignoring process constraints such as low ergonomics, monotony or high effort.
The technological challenges of the project are the following:
- Development of an artificial vision system based on simple and cheap RGB cameras and image processing using convolutional neural networks and semantic segmentation.
- Development of a simultaneous location and mapping system based on Visual SLAM architecture algorithms for application in the industrial field.
- Development of an optimal path planning system (Path Planning) that allows the mobile robot to autonomously define the optimal path to the target based on the data handled by the vehicle in real time, and the data obtained in situations experienced in the past (machine learning).
- Development of an intelligent navigation system (Path following) that allows the mobile robot to react to obstacles and unforeseen events, and navigate safely and uninterruptedly towards the objective in environments with a high density of traffic and obstacles.
- Vision manipulation system development.
- Development of a synchronism system between the movement of the robot and the manipulator arm for dynamic handling operations.
- Electromechanical integration in the mobile robot of a handling system based on a robotic articulated arm with precise movement control and high maneuverability.
Argolabe's roleArgolabe Ingeniería, S.L. is in charge of managing the project, carrying out and coordinating the development work with the Universityd
Project lead partner:
Partners and collaborators: