Key Takeaways
-
Sub-30 cm Precision Without Tags: Real-life results show that markerless indirect tracking achieves sub-30 cm accuracy indoors and sub-10 cm outdoors, matching the performance of traditional direct-tagging methods.
-
Unlimited Scalability: By tracking the material handling equipment (MHE) rather than individual items, the system allows for an unlimited number of tracked payloads without additional hardware, battery maintenance, or per-item costs.
-
Fully Automated Future Logistics: In the future, the system can detect pick-up and drop-down events automatically using RFID or other identification technologies, ensuring 3D inventory visibility without requiring manual input from forklift operators.
Markerless Indirect Tracking
Today, industry and warehousing are rapidly integrating intelligent technologies to optimise warehouse management processes. This aims to improve overall efficiency and transparency while reducing the impact of human error. A critical component of this optimisation is the real-time localisation of machinery, semi-finished products, and materials. Accurate tracking minimises operational bottlenecks, such as delays and misplacements.
However, standard direct positioning technologies face limitations in certain industrial environments. Materials undergoing frequent mechanical processing cannot have tags or trackers attached to it, making direct localisation impossible. To address these needs, we developed a proof-of-concept for a markerless tracking solution for the automatic, three-dimensional indirect localization of industrial products without direct tagging.
Technical Criteria for Indirect Tracking
Our proof-of-concept solution is based on the following requirements :
-
Markerless Operation: Tracked products and equipment must remain fully unmarked by any tracking units.
-
3D Support: The system must provide automatic three-dimensional tracking to cover real-life industrial scenarios like shelving and stacking.
-
Minimal Integration: Setup components should require minimal integration into existing industrial machinery mechanisms.
-
Cost-Efficiency: The system must minimise the number of extra supporting sensors and hardware units.
The Indirect Tracking
In this approach, payloads are tracked indirectly during transportation by material handling equipment (MHE), such as forklifts. The location of the payload is directly associated with the location of the MHE’s handling unit, specifically the forks .
The system architecture relies on a set of sensors deployed exclusively on the forklift and processed on the server side:
-
Horizontal Location: The real-time location of the MHE is provided by an underlying positioning system – UWB RTLS indoors and GNSS outdoors. Coordinates of the two systems are integrated in the Eliko software.
-
Two-Dimensional Estimation: The MHE’s 2D location is estimated using its known location and heading information.
-
Elevation Monitoring: Real-time fork elevation is monitored by a separate dedicated unit.
-
Automated Detection: The method included a semi-automatic detection of pick-up and drop-down events, which leads the way for a fully automatic pallet detection system in the future
Proof-of-Concept and Results
The performance of this markerless tracking method was verified through a proof-of-concept deployed on a full-scale Linde h20 forklift in indoor and outdoor environments. The test was conducted in an active industrial woodworking facility (Auroom Wellness) during standard material transportation routines.
Indoor Test Parameters
The test scenarios involved transporting two pallets (1 x 1.5 m) between multiple storage spots. The scenarios covered:
-
Standard indoor transportation.
-
Shelving and stacking operations.
-
Stacked transportation scenarios.
Advantages of a Markerless Indoor Tracking
The transition from direct to markerless indirect tracking provides several benefits for industrial scalability and maintenance:
Since tracked payloads do not require direct marking, the system offers good scalability. There is no additional per unit cost as the volume of tracked materials increases.
The reliable detection of pick-up and drop-down events enables fully automatic tracking in the future. This ensures the forklift operator is not directly involved in the tracking process, reducing the probability of human errors and operational delays.
Conclusion
Our test results show that markerless indirect tracking allows for the automatic 3D localisation of industrial products in an accurate and reliable way. By combining Eliko RTLS UWB tracking with GNSS RTK, it’s possible to offer consistent 10-30 cm tracking combined with pallet height measurement.
By equipping the material handling unit with a tracking system and not each individual pallet, manufacturers can achieve high-precision visibility for detecting bottlenecks and improving warehouse operations.
