Precision Fusion Positioning System for Overhead Cranes
Kelude's Heavy-Load High-Precision Fusion Positioning System for Overhead Cranes Passes Steel Mill Acceptance with ±5mm Accuracy. Kelude Heavy Industry's self-developed fusion positioning system completed acceptance testing on a 50t casting overhead crane at a steel mill, consistently achieving ±5mm positioning accuracy under full-load conditions—meeting the precise positioning requirements for unmanned overhead crane automatic lifting and transport.
Kelude Heavy Industry's self-developed heavy-load high-precision fusion positioning system for overhead cranes has passed acceptance testing on a 50t casting crane at a steel mill, delivering stable ±5mm positioning accuracy under full-load conditions and satisfying the precise positioning demands of unmanned automatic crane operations. The system integrates three independent positioning sources—LiDAR, absolute encoders, and ultra-wideband (UWB)—fused in real time through an Extended Kalman Filter (EKF) algorithm, filling a critical gap in domestically produced cranes for heavy-load, high-precision positioning applications.
Three-Source Fusion Positioning Technology
The system employs three independent positioning units—LiDAR, absolute encoders, and UWB—each generating positioning data that is fed into an EKF fusion engine to compute the optimal position estimate. The LiDAR unit uses a SICK LMS111 indoor single-line scanning radar mounted on the underside of the trolley, scanning ground feature points with a planar measurement accuracy of ±12mm, a 270° scanning angle, and a 50Hz data refresh rate. Absolute encoders are installed on the motor shafts of both the crane bridge and trolley drive motors, measuring travel displacement through a rack-and-pinion transmission with 0.1mm resolution; cumulative error is controlled to within ±3mm after end-position calibration. UWB positioning base stations are deployed every 15 meters along the crane rail, with tags mounted on the side of the trolley frame, providing ranging accuracy of ±10cm and serving as a global position reference to eliminate cumulative drift from the LiDAR and encoders. The EKF fusion engine runs on a Beckhoff CX2030 embedded controller with a 4ms computation cycle. The three position data streams undergo time alignment, outlier rejection, and Kalman gain weighting to output the optimal position estimate, achieving a measured positioning accuracy of ±5mm (3σ) under full-load operation of the 50t casting crane.
| Comparison Item | Li DAR | Absolute encoder | UWB positioning | Triple-Source Fusion |
|---|---|---|---|---|
| Single-Source Accuracy | ±12mm | ±3mm(Calibration After) | ±10cm | ±5mm |
| Defect | High Susceptibility to Dust Obscuration | Cumulative Error Requires Calibration | Accuracy Low, Multipath Interference | Mutual Compensation |
| Refresh Rate Frequency | 50Hz | Continuous | 10Hz | 250Hz |
| environmental adaptability | Medium(Dust-Sensitive) | High | Medium(Metal-Obstruction Sensitive) | High |
Steel Mill Acceptance Test Results
Acceptance testing was conducted on a 50t casting overhead crane in the steelmaking workshop of a steel mill, under ambient temperatures of 45–65°C, dust concentrations of approximately 8 mg/m³, and strong electromagnetic interference from VFDs and high-power electric arc furnaces. The test program followed the requirements of ISO 4301 Crane Design Standard and FEM 1.001 Crane Test Specification, covering four core indicators: static positioning accuracy, dynamic tracking accuracy, repeat positioning accuracy, and multi-source failure switching. Test results showed static positioning accuracy of ±4.2 mm (average of 50 measurements), dynamic tracking accuracy of ±5.8 mm (with the trolley running at 0.5 m/s), repeat positioning accuracy of ±3.1 mm (10 repeated tests at the same position), and multi-source failure switching time of less than 200 ms (when any single sensor signal is lost, the fusion system automatically switches to the remaining sensors and continues outputting position data). All three key accuracy indicators outperformed the design targets.
Deployment Across Multiple Industries
Prior to passing the steel mill acceptance test, the positioning system had already been deployed and validated across 8 projects in 5 industries, covering steel and metallurgy (3 casting crane projects), heavy machinery manufacturing (2 projects), non-ferrous metals (1 project), automotive manufacturing (1 project), and papermaking (1 project). The deployed cranes range from 10t to 100t capacity, with work duty classifications of A5 to A7 and rail spans from 12 m to 31.5 m. Of the 8 projects, 6 achieved fully automatic lifting and transport operations with unmanned overhead cranes, while 2 adopted semi-automatic operation modes with operator assistance. The average on-site commissioning period for the positioning system was 5–8 working days, approximately 50% shorter than the 14 working days required during the initial development phase. The system has accumulated over 12,000 hours of failure-free operation, with no equipment collisions or lifting accidents caused by positioning failures. The sensor fusion algorithm and fault-tolerant design comply with the safety monitoring and management system requirements of ISO 4301, and the system supports data interfacing with Kelude's MCSS multi-crane dispatching system, providing high-precision position data for unmanned crane scheduling in smart factories.
Related technical resources: Overhead Crane High-Precision Positioning Technology Comparison: LiDAR, Encoder, and UWB Fusion Selection Guide (a detailed breakdown of six positioning approaches, covering technical principles, cost comparison, and selection workflow). AGV/RGV and Overhead Crane Multi-Vehicle Coordination: Smart Factory Unmanned Transport Dispatching in Practice (integration of the MCSS multi-crane dispatching system with the positioning system).
FAQ: Positioning System Performance and Retrofit
Q: Can the three-source fusion positioning system operate reliably under strong electromagnetic interference?
A: Yes. The system's reliability in strong electromagnetic environments was verified during the steel mill acceptance test. The extended Kalman filter fusion mechanism automatically reduces the fusion weight of any sensor whose signal is affected by interference, while the remaining sensors continue to provide position data. In a full-load test on a 50t casting crane, the system achieved ±5 mm positioning accuracy in an electromagnetic environment where VFDs and high-power arc furnaces were operating simultaneously. All sensors and communication links use industrial-grade shielded cables and metal enclosures, meeting the electromagnetic compatibility requirements for electrical equipment specified in ISO 4301.
Q: How is the retrofit designed for existing overhead cranes?
A: The system is engineered with retrofit scenarios in mind, using a modular deployment approach. The LiDAR sensor mounts to a pre-drilled bracket on the underside of the trolley, while the absolute encoder requires a coupling and encoder housing to be added at the crane drive end. UWB anchors are installed along the crane rail sidewalls or on columns. A typical retrofit takes 3–5 working days (excluding system commissioning), and the renovation cost covers sensor hardware, the embedded controller, and commissioning services. For cranes already equipped with Variable Frequency Drives (VFD) and PLC control, the system interfaces with the existing control cabinet via PROFINET or EtherCAT bus—no replacement of the electrical control system is required.
Q: Can the positioning system interface with the customer's MES or WMS?
A: Yes. Position data is published through an OPC UA server, with a data model covering 16 data points including crane bridge position, trolley position, positioning confidence, and sensor health status. The customer's MES or Warehouse Management System (WMS) can subscribe to the position data via an OPC UA client to enable automatic dispatch of lifting and transport tasks and confirmation of job completion. The system has been successfully integrated with MES platforms in three projects, with the data refresh cycle synchronized to the positioning system output (4 ms) and end-to-end communication latency not exceeding 50 ms.
Q: Do multiple positioning systems interfere with each other when several cranes operate on the same bay?
A: In multi-crane operation on the same bay, the LiDAR and encoders operate independently with no mutual interference. UWB positioning uses Time Division Multiple Access (TDMA) to allocate time slots, with each crane's UWB tag transmitting in a different time slot to prevent signal collision. The system is designed to support up to 8 cranes operating simultaneously on the same bay without interference, and positioning reliability under multi-crane concurrent operation has been verified in two steel mill projects. UWB anchor installation height and angle are optimized through a site survey to ensure each crane's tag is covered by at least 3 anchors along the full length of the crane rail.
