Forklift Load Detection Using AI Vision for Center-of-Gravity Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for load detection in forklift trucks are inadequate for handling complexly structured loads, leading to safety issues, inefficiencies, and increased costs due to the need for manual or stationary contour checks, and limitations in determining load dimensions, weight distribution, and center of gravity.
Innovation Solution
Utilizing optical sensors, such as cameras, mounted on forklifts or drones, combined with data processing units employing artificial intelligence to evaluate sensor data for precise load detection, including dimensions, weight, and center of gravity, enabling flexible and automated load handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If stationary contour inspection stations are used to detect load overhang, then load safety can be ensured, but system complexity and transport time increase due to additional check stations and complex transport processes
Solution Approach 1:
The patent replaces mechanical/optical inspection systems with electromagnetic sensing. Load cells integrated into the forks use strain gauge technology to detect load weight and center of gravity, while RFID readers use electromagnetic fields to identify load characteristics. This substitution eliminates the need for complex mechanical contour inspection stations while maintaining safety.
Solution Approach 2:
The forklift's existing components are made multi-functional. The forks not only lift loads but also contain integrated load cells for weight detection. The control unit not only manages basic operations but also processes sensor data to determine load center of gravity and adjust driving parameters. This multi-functionality eliminates dedicated inspection stations.
2Loss of information
If extensive cabling is used at the loading platform for binary load detection, then load detection capability is provided, but fire hazards increase due to cables on moving parts
Solution Approach 1:
The patent replaces physical cable connections with wireless communication systems. RFID readers and load cells transmit data wirelessly to the control unit using electromagnetic signals, eliminating cables from the loading platform and moving parts, thus removing the fire hazard while maintaining full load detection capability.
Solution Approach 2:
The patent introduces wireless communication as an intermediary between sensors and the control unit. Instead of direct cable connections, electromagnetic waves serve as the medium for data transmission, allowing load information to be conveyed without physical cables that could pose fire risks.
3Reliability
If the worst-case scenario is assumed for unknown load characteristics, then safety is maintained, but performance decreases due to limited speed and turning angle
Solution Approach 1:
The patent implements real-time feedback through integrated sensors that continuously monitor load weight, center of gravity, and distribution. The control unit receives this feedback and dynamically adjusts driving speed, turning angle, and lifting operations accordingly. This allows the system to operate at optimal performance levels matched to actual load conditions rather than conservative worst-case assumptions.
Solution Approach 2:
The patent makes the forklift's operational parameters dynamic rather than static. The control unit continuously adapts speed limits, turning radii, and lifting speeds based on real-time load characteristics detected by sensors. This dynamic adjustment enables high performance when loads are favorable while maintaining safety when loads are challenging.
4Reliability
If manual load contour checks are performed, then load safety can be verified, but productivity decreases due to time-consuming manual intervention
Solution Approach 1:
The patent enables the forklift to perform self-inspection of load characteristics through integrated sensors and automated control systems. The load cells automatically detect weight and center of gravity, RFID readers identify load types, and the control unit automatically determines safety parameters. This self-service capability eliminates the need for manual load contour checks while maintaining safety verification.
Solution Approach 2:
The patent replaces manual visual inspection with automated electromagnetic sensing. Load cells use strain measurement technology to detect weight, RFID systems use electromagnetic fields to identify loads, and optical sensors automatically scan load contours. This automation eliminates time-consuming manual checks while providing more precise and consistent safety verification.
Data Source
Figure 1
Figure 2A~2D
Figure 3~4
AI summary
The invention relates to a method for load detection in an industrial truck (2, 3, 4) and to a system (1) for carrying out the method. It is proposed that the load (L) is detected by means of at least one optical sensor (S) and identified by evaluating the sensor data by means of artificial intelligence in at least one data processing unit (D).