エウレカロボティックス Product Lineup

This is a case where Kawasaki Heavy Industries' semi-humanoid robot "Nyokkey" has adopted our "Eureka AI Vision System." Translucent materials transmit and scatter light, making it difficult to capture shapes with projection-based 3D cameras. Additionally, there were variations in illumination caused by spotlights in the environment. This system employs a stereo camera approach. Since it does not depend on projection patterns, it enables stable three-dimensional measurement even in environments with translucent materials or varying illumination. As a result, the three-dimensional position and posture of translucent workpieces were calculated with high precision, achieving stable picking. [Case Overview] ■ Challenges - Translucent materials make shape acquisition difficult with projection-based 3D cameras. - Variations in illumination caused by spotlights also occur. ■ Reasons for Adoption - Presented effective solutions in collaboration with Shōtoku Seisakusho in response to complex technical requirements. - The company's commitment to tackling the issue with full force and strong dedication to customer requests was highly evaluated. *For more details, please refer to the related links or feel free to contact us.

• Image analysis

We would like to introduce a case where our "Eureka AI Vision System" has been adopted as a standard feature in Aioi System's general-purpose automatic picking device. In logistics sites, automation investments are accelerating in processes such as receiving, storage, transportation, and sorting, against the backdrop of a serious labor shortage. On the other hand, the final picking process before shipment often relies heavily on human labor due to the need to handle a wide variety of products with high mixed loads and frequently changing SKU counts, resulting in a significant operational burden. With this adoption, Aioi System is advancing the establishment of a system that can consistently propose solutions from upstream to downstream in the logistics process, including the final picking process, which has been considered difficult to automate until now. [Case Overview] ■ Background: The "final barrier" of logistics automation ■ Problem Resolution: General-purpose automation that "sees and picks immediately" ■ System Mechanism: Adoption of a masterless picking method ■ Why Eureka: AI vision that fits logistics sites *For more details, please refer to the related links or feel free to contact us.

• Image analysis

Bulk picking has made significant progress over the past decade, but there remains a considerable gap between "moving systems" and "systems that continue to operate stably in factories." In fact, common "oversights" are present in sites where projects are struggling. - Assuming a "100% success rate" - Underestimating the impact of "workpiece oil and reflections" on imaging - Not accounting for the increase in cycle time due to "3D camera placement" Such "realities faced on-site" are rarely discussed in advance by vendors. At Eureka Robotics, we have condensed insights gained from over 30 million mass production picks into a guidebook that outlines essential points to confirm before implementation. ■ What you can learn from this document: System configuration for bulk picking Characteristics of workpieces that are likely to succeed or fail We hope you find this useful in leading your automation project to success.

• Image Processing Software

We would like to introduce an example of the automation of picking and inspection of side mirror covers at our company. The rising labor costs, the enormous time required for inspection, and the challenge of automating cases where the position and orientation of the workpieces vary were significant issues. The accuracy depended on the experience and condition of the operators, making it difficult to visually detect minute scratches and dents. After implementation, we were able to automatically identify fine scratches and dents with high precision. We have seamlessly integrated the entire workflow from bulk picking to visual inspection, enhancing quality assurance. [Challenges] ■ Limitations of manual inspection: Inspection accuracy depends on the experience and condition of the operators ■ Overlooking minor defects: It is difficult to visually detect minute scratches and dents ■ Challenges with cost and productivity: Rising labor costs and the enormous time required for inspection ■ Responding to bulk picking: Automation is difficult when the position and orientation of the workpieces vary *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

We would like to introduce a case where automatic fitting of connectors flowing on a line, which has been considered difficult to automate until now, has been realized. Fitting requires strong pressure, and delicate tactile feedback was necessary to ensure it was done correctly. It was challenging for the robot to accurately detect the position and orientation for gripping. In this system, the robot accurately recognizes the orientation of the connector on the slider using a 3D camera and AI technology. It receives real-time feedback from force sensors to delicately guide the insertion. 【Case Overview】 ■Challenges - When the target object is not stationary, it is difficult for the robot to accurately detect and grip its position and orientation. - Fitting requires strong pressure, and delicate "tactile feedback" is necessary to ensure it is done correctly. ■Results - Automation of manual processes: Achieved automation of high-precision tasks that relied on human labor. - Improved productivity and consistency: Automation ensures stable productivity. - Complete automation: Integrated conveyor tracking and high-precision insertion into a single seamless robot workflow. *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

We would like to introduce a case study of the implementation of the "Eureka AI Vision System" at Sigma Optical, a manufacturer of optical equipment. In lens inspection, there is a need to accommodate various sizes and shapes, and manual inspection required an enormous amount of effort. A solution that achieves accurate defect detection while maintaining high cost-effectiveness was necessary. With the introduction of this system, deep learning was utilized to detect minute defects, improving inspection accuracy. It eliminated manual inspections and enabled faster lens inspections without compromising precision. 【Case Overview】 ■Challenges - Lens inspection requires accommodating various sizes and shapes. - Manual inspections carry a high risk of human error and require substantial effort. ■Results - Quality Improvement: Achieved a detection success rate close to 100%, ensuring high-quality lens products. - Cost Reduction: Eliminated manual inspections and improved operational efficiency, reducing labor costs. *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

We would like to introduce an automation case utilizing the "Eureka AI Vision System" at Bridgestone Soft Robotics. In e-commerce, it is unpredictable what kind of orders will come in, and dealing with a variety of products and a wide range of SKUs has been a challenge. Additionally, conventional 3D vision systems have difficulty recognizing glossy and transparent objects. With the introduction of this system, it has become possible to pick objects without the need for prior data registration or learning. The system automatically selects suction pads and claws according to the characteristics of the objects, achieving high reliability and fast processing. 【Case Overview】 ■ Challenges - Responding to diverse products: size, type, shape - A wide variety of SKUs: unpredictable orders in e-commerce - Objects that are difficult to recognize with conventional 3D vision systems: glossy, thin, transparent/semi-transparent ■ Results - Demonstrated automation of piece picking - Automatically selected appropriate gripping methods - Achieved high reliability and fast processing *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can achieve automatic weighing and balancing of engine blades. By simply setting the tray and pressing the start button, we eliminate the possibility of recording and measurement errors, achieving zero human error. This frees operators from repetitive tasks, allowing them to transition to more value-added work. By scanning the serial number/part number, measurement results are automatically recorded. The blades are repositioned to achieve the appropriate balance, and a report is automatically generated, including final placement, imbalance amount, and complete traceability. 【Features】 ■ Fully automated weighing, inspection, and balancing of up to 70 blades in 25 minutes ■ Scans serial numbers/part numbers and automatically records measurement results ■ Automatically optimizes blade placement *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

System integrators utilize the Eureka AI vision system to build robotic systems more quickly and cost-effectively. By integrating vision into robotic cells with standardized tools and workflows, we minimize adjustment labor and rework. Our team, with a proven track record of implementations worldwide, provides practical and supportive assistance. We support the success of all projects with a stable vision platform that avoids issues caused by unstable vision configurations. 【Features】 ■ Set up in half a day and start picking immediately ■ Changes to processes and work can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can automate the injection molding process using AI vision technology. We can position and place inserts and molded parts into cavities and fixtures with high reproducibility, reducing dependence on bowl feeders and manual setups. We verify the presence, orientation, and surface quality of molded parts for each cycle, detecting defects before they flow into packaging or assembly processes. We automatically pick molded parts from containers or conveyors and sort them into good and defective products. 【Features】 ■ Set up in half a day and start picking immediately ■ Process and work changes can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can automate the sorting of workpieces using AI vision. We can detect, locate, and classify multiple types of workpieces mixed together on a single container or conveyor without prior sorting. During the picking operation, we check for the presence, orientation, and quality of the workpieces, sorting good and defective items on the spot. We can handle variations in shape and disorganized conditions, replacing manual sorting tasks. We will ensure the automation of sorting tasks that were previously done by hand. 【Features】 ■ Set up in half a day and start picking immediately ■ Changes in processes and workpieces can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can automate assembly tasks using AI vision. We detect the workpiece in 3D and automatically align it to the correct position and orientation before insertion. It is robust against variations in workpieces and jigs, achieving stable assembly. For the workpieces after placement, we automatically check for presence, orientation, and correct seating status, detecting issues early to prevent impacts on subsequent processes. 【Features】 ■ Set up in half a day and start picking immediately ■ Changes in processes or workpieces can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can automate the loading and unloading of workpieces using AI vision. It can detect and locate workpieces in bulk state, eliminating the need for high-precision jigs or pre-alignment. The robot guides the workpieces into the correct position when loading into CNC machines, lathes, and presses, reducing the risk of jams, interference, collisions, and defects caused by positional misalignment. It corrects for part tolerances, tray positional deviations, and jig variations, achieving stable operation unaffected by changing conditions. It accommodates variations in workpieces without the need for re-teaching. 【Features】 ■ Set up in half a day and start picking immediately ■ Process and work changes can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera

Our company can achieve 3D random bin picking with AI vision. We integrate 3D cameras, controllers, and software. Even without specialized knowledge, you can quickly and easily set up robot vision. We have conducted over 25 million picks, and companies of various sizes, from large corporations to small and medium-sized enterprises, have adopted our solution. We provide reliable 3D random bin picking. 【Features】 ■ Set up in half a day and start picking immediately ■ Process and work changes can be completed on-site ■ Compatible with major robots and PLCs *For more details, please download the PDF or feel free to contact us.

• High Speed Camera