Low Cost Scanning Device Application for Footwear Industry

Notwithstanding laser scanning technology is a mature technology used in widening application fields, a variety of barriers hinds its integration in robotized production lines. To face well know problems, as high-costs and not customizable solutions among others, the authors have developed a new family of laser scanning devices based on off the shelves and low cost hardware components and on a modular design that allow the customization of the device according to the specific application requirements. High accuracy is guaranteed through a computationally efficient non-parametric calibration procedure. Keeping limited the overall cost of the solution provided while increasing in general the sensing capabilities of an Industrial Robot (IR) and in particular the autonomous recognition of position, orientation and furthermore shape and geometrical features of objects within the robotic workspace, can boost the penetration of IRs in typical traditional industrial sectors where SMEs productive scenario is mainly characterized by manual fabrication processes, high product variability, small batches and little capital investments. High added value footwear industry is a paradigmatic example where hard automation is limited from one side by high variability of products and huge request of autonomous adaptation to cope with different loosely structured fabrication processes and from the other side by a low propensity towards high capital investment that make difficult the penetration of industrial robots. Within the framework of the European Project ROBOFOOT (EU-FP7-SMP), authors have conceived and developed a new family of low cost modular and reconfigurable laser scanners and successfully applied in the footwear fabrication scenario. Various operations should be improved by actual measure of the shoe being manufactured, and among the others, the identification of the relative positioning between the last (the plastic element on top of which is built the shoe) and the gripping device is extremely critical because it is still performed manually. Positioning errors in this phase are critical since all robot part-programs depend on the correct alignment of the last with the robot end-effector. To face this problem, a procedure, based on the Iterative Closest Points (ICP) optimization method, has been developed and integrated on board of the laser scanning device that directly communicates with the robot controller to adapt and autonomously correct the part-program to align the tool nominal path with the actual shoe being manufactured. Efficacy of the proposed methods has been proved by measuring the interaction forces between the tool and the last handled by a robot during different technological operations (roughing, polishing etc) typical of the shoe fabrication cycle.