Collaborative robots + AI + machine vision boost agricultural and manufacturing capabilities
Author : Yvonne Zhang, Interact Analysis
05 March 2024
GRIT cobot solution from Kane Robotics
The combination of AI and machine vision is generating more practical applications for collaborative robots (cobots), particularly in modern agricultural and manufacturing scenarios.
Growth of AI + machine vision in cobots
The application scenarios of AI + machine vision in collaborative robots are gradually expanding, with increasing penetration rates. Machine vision can assist collaborative robots in more accurately identifying and tracking targets. Combined with artificial intelligence decision-making capabilities, collaborative robots can quickly learn and optimise methods for task execution, achieving higher efficiency in task completion.
Here are several practical examples:
Denso Robotics showcased its new collaborative robot, Cobotta Pro, along with a vision system for scooter assembly at the 2023 iREX Exhibition. This demonstration highlighted the advantages of integrating artificial intelligence and vision systems into collaborative robots: they can read QR codes, perform intelligent position correction, and recognise human
commands through a voice-controlled IPC, allowing flexible switching of assembly steps. Collaborative robots can accurately grip the frame and work in coordination with workers to assemble tires and handlebars.
Kane Robotics from the United States has also combined artificial intelligence with machine vision, enabling its collaborative robots to automatically track and polish weld seams with high precision and speed, showcasing the potential of AI in fine operations.
Doosan Robotics and AiV, a South Korean industrial deep learning computer vision technology company, jointly introduced the new Otto Matic palletising system, which applies a combination of AI + machine vision + collaborative robots to the palletising process. This system can handle unstructured and randomly sized boxes to improve the efficiency of logistics automation.
Yvonne Zhang, Research Associate, Interact Analysis
AI + Machine Vision + Cobots for modern agricultural applications
With the advancement of technology, collaborative robots equipped with vision systems and artificial intelligence are gradually changing traditional agricultural practices, and are being applied to various agricultural picking applications.
In 2023, research teams from the Netherlands and Switzerland successfully created a tomato-picking robot, pairing generative artificial intelligence ChatGPT with a machine vision system. This robot captures images through cameras, and utilises ChatGPT for image recognition. Meanwhile, ChatGPT can communicate in real-time with researchers, asking questions about tomato ripeness, picking techniques, and more, making decisions based on the information received. This highly-intelligent collaborative approach enables the cobot to identify and pick ripe tomatoes accurately, introducing new picking capabilities to modern agriculture.
In addition to research teams, collaborative robot companies have also begun to explore the integration of AI, machine vision, and collaborative robot technologies in practical applications. For example, Flexiv Robotics has applied these technologies to cabbage harvesting.
Dobot’s Nova collaborative robot has been successfully deployed for strawberry picking. It utilises AI + vision technology to accurately identify the ripeness of strawberries, and combines it with an AGV (automated guided vehicle) for efficient navigation in the field. This integration has significantly boosted the efficiency of traditional manual picking.
Interact Analysis has observed a gradual increase in the application of collaborative robots in the agricultural sector. To reflect this trend, we have added an agricultural category to the sixth edition of the collaborative robot report, which is set to be released in April this year.
More information on this report here.
Contact Details and Archive...