SkyVision Enhances Production Rhythm and Full Inspection Capacity for Two - hand Operations

Wemio's SkyVision 0-code video surveillance AI platform (with features such as on - site hourly training of self-owned models, behavior/event recognition, real-time alerts from edge boxes, 100% local data storage without leaving the site, and semantic understanding of the DaoAI World model) has improved the production rhythm by 30% and significantly increased the 100% full-inspection capacity through precise comparison of two-hand operation specifications.

In the industrial SOP/operation specification industry, the strict implementation of two-hand operation specifications is crucial for ensuring production quality and efficiency. For example, in the production line of an auto parts manufacturing enterprise, workers need to assemble parts with both hands in a coordinated manner. During this process, the degree of compliance of two-hand operations directly affects product quality and production rhythm. Traditional monitoring methods struggle to meet the requirements of high-capacity and 100% full inspection. The introduction of Wemio's SkyVision 0-code video surveillance AI platform provides an effective solution to this problem. Through precise comparison of two-hand operation specifications, it has successfully improved the production rhythm and 100% full-inspection capacity.

Pain Points: Why is it So Difficult?

From the perspective of quantitative difficulties, first, the false-negative rate is relatively high. Under traditional monitoring methods, the false-negative rate can reach about 10%. This means that some non-compliant two-hand operations may be overlooked, affecting product quality. Second, the false-positive rate cannot be underestimated, approximately 8%. Frequent false alarms not only increase the man-hours for manual re-judgment but also distract operators. In terms of manual re-judgment man-hours, due to the need to manually confirm a large number of suspected non-compliant situations, the manual re-judgment man-hours per production line can reach more than 4 hours per day. Moreover, the compliance risk of the production line is relatively high. Once there are non-compliant two-hand operations, it may lead to product quality problems and safety accidents. Finally, the unit cost also increases due to these problems. It is estimated that the cost per product increases by about 5%.

Analyzing from the root-cause level, in terms of process, the actions of two-hand operations are complex and diverse, and the operating habits of different workers also vary. This makes it difficult for traditional rule-based AOI to accurately define all standard actions. In terms of imaging, the unstable lighting conditions in the workshop may cause image blurring, affecting the accuracy of detection. From the material perspective, the different shapes and colors of parts also increase the difficulty of recognition. The requirement of production rhythm means that the detection must be fast and efficient, and traditional methods struggle to balance speed and accuracy. Combining with the current hot topic of the digital and domestic development path of campus security under the '14th Five - Year Plan for Education Development', the industrial field also urgently needs digital solutions to improve production efficiency and quality.

Technical Principle

Wemio's SkyVision 0-code video surveillance AI platform uses advanced deep-learning algorithms. Through on - site hourly training of self-owned models, it can quickly adapt to different two-hand operation scenarios. The platform uses the DaoAI World model for semantic understanding to accurately identify the actions and positions of both hands. In terms of hardware, it is equipped with high-performance edge boxes that can process video data in real-time and issue alerts. The combination of this algorithm and hardware enables the platform to accurately determine whether two-hand operations are compliant. Compared with traditional rule-based AOI, rule-based AOI makes judgments based on preset rules and has poor adaptability to complex and changeable two-hand operation scenarios. SkyVision can continuously learn and optimize through deep learning, improving the accuracy of detection. Compared with manual visual inspection, manual visual inspection has problems of fatigue and subjective judgment, while SkyVision can achieve 24-hour uninterrupted, objective, and accurate detection, reducing the false-negative rate to < 1%.

In addition, the 100% local data storage feature of the platform ensures the security and privacy of data. During the data-processing process, all data is analyzed and processed locally without being uploaded to the cloud, avoiding the risk of data leakage. At the same time, the cross-scenario generalization ability of the DaoAI World model enables the platform to be quickly applied to different industrial scenarios, improving the platform's versatility and applicability.

Typical Application Scenarios

• Two - hand parts assembly process: The monitoring camera captures the actions of both hands in real-time to detect whether the assembly is carried out in the specified order and with the specified force. The difficulty lies in the large differences in assembly actions for different parts, and the platform needs to quickly adapt and accurately identify. Wemio's SkyVision can quickly master the specifications of different assembly processes through on - site model training.
• Two - hand welding process: It detects whether the position, angle, and welding time of both hands during the welding process meet the standards. Since strong light and smoke are generated during the welding process, affecting the clarity of the image, this is the difficulty of detection. SkyVision's advanced algorithm can effectively filter interference and accurately judge whether the welding action is compliant.
• Two - hand material handling process: It judges whether the posture and force of both hands when handling materials are correct to prevent material damage and personal injury. The different shapes and weights of materials increase the difficulty of detection. The platform can accurately identify the handling specifications of different materials through semantic understanding.
• Two - hand equipment operation process: It monitors whether both hands operate the equipment in accordance with the operating procedures to avoid equipment damage and safety accidents caused by misoperation. The operation interfaces and actions of equipment are complex and diverse. SkyVision can accurately judge whether the operation is compliant through learning and analysis.

Implementation Case

A medium-sized auto parts manufacturing enterprise with multiple two-hand operation production lines. Before introducing Wemio's SkyVision 0-code video surveillance AI platform, the false-negative rate of the production line was 10%, the false-positive rate was 8%, the manual re-judgment man-hours per production line per day exceeded 4 hours, the production rhythm was slow, and it was difficult to meet the demand for 100% full inspection. During the implementation process, Wemio's technical team first conducted a detailed review and analysis of the two-hand operation specifications of the production line, and then used the SkyVision platform for on - site hourly training of self-owned models. After a period of debugging and optimization, the platform was successfully launched. After the launch, the false-negative rate was reduced to < 1%, the false-positive rate was reduced by -7%, the manual re-judgment man-hours per production line per day were reduced to less than 1 hour, the production rhythm was increased by 30%, and the 100% full-inspection capacity was significantly improved.

Wemio's SkyVision platform has brought a qualitative leap to the detection of industrial two-hand operation specifications, effectively improving production line efficiency and quality.

Wemio's Solution and Product

Wemio takes the SkyVision 0-code video surveillance AI platform as the core and provides a comprehensive solution. In terms of modeling, through on - site hourly training of self-owned models, it can quickly adapt to the two-hand operation specifications of different customers. In terms of model change, due to the use of 0-code programming, the model-change time can be controlled within 5 minutes, greatly improving the flexibility of the production line. In terms of deployment, it supports 100% local private deployment, and the data does not leave the site, ensuring data security. At the same time, the platform can be integrated with the customer's existing systems through SDK/API/Docker, etc. In addition, Wemio can also provide supporting DaoAI AI AOI software systems and DaoAI 2D/3D AI AOI equipment according to customer needs to further improve the accuracy and efficiency of detection.

By using Wemio's SkyVision platform, the enterprise has achieved significant business value. In terms of indicators, the false-negative rate has been reduced to < 1%, the false-positive rate has been reduced by -7%, and the production rhythm has been increased by 30%. This not only improves product quality, reduces the cost increase caused by quality problems, but also improves production efficiency and reduces labor costs. At the same time, the improvement of 100% full-inspection capacity enables the enterprise to better meet market demand and enhance its competitiveness.

This article was generated by AI. Customer cases are simulated scenarios based on real product capabilities and figures are illustrative; see product pages for official benchmarks.