a b s t r a c  t

Keywords: Robot Machining Small-batch Industrial HMI Compliance control Path planning

Hard materials can be cost effectively machined with standard industrial robots by enhancing current state-of- the-art technologies. It is demonstrated that even hard metals with specific robotics-optimised novel hard-metal tools can be machined by standard industrial robots with an improved position-control approach and enhanced compliance-control functions. It also shows that the novel strategies to compensate for elastic robot errors, based on models and advanced control, as well as the utilisation of new affordable sensors and human-machine in- terfaces, can considerably improve the robot performance and applicability of robots in machining tasks. In conjunction with the development of safe robots for human-robot collaboration and cooperation, the results of this paper provide a solid background for establishing industrial robots for industrial-machining applications in both small- and medium-size enterprises and large industry. The planned short-term and long-term exploitation of the results should significantly increase the future robot usage in the machining operations.

1. Introduction

Hard-materials machining has attracted a large amount of atten- tion from advanced industries, in particular, European automotive, aerospace and biomedical industries. However, existing technology has failed to provide these industries with a cost-efficient solution to accom- modate small-batch production of large and complex-shaped products. Until now, only about 3% of overall industrial robots employed in industry are used for machining. This is quite in opposition to the mar- ket potential and benefits of industrial-robot-machining applications. It has been widely recognised that inherent 5+ axis machining capabil- ity combined with flexibility, large-work envelope and multiple-station capability is a flexible solution that allows users to expand the range of machining applications at a price point competitive to that of employ- ing a traditional CNC (Computer Numeric Control) machine. With the recent dynamic-cost reduction and performance optimisation of mod- ern industrial robots, the price of a comparable robotic solution is typi- cally 1/5-1/3 of the cost of a CNC machine. Integration of two or more robots into flexible multi-stationed and multi-tooled robotic machining cells may result in significantly lower-cost investments in comparison to employing large-CNC machines.

Based on several studies [1,2], the two underlying technical limi- tations for the widespread adoption of robotic machining are the in- sufficient robustness of robotic structures (insufficient precision and stiffness) and lack of efficient programming tools that transfer CAD (Computer-aided design) models into robot motion. Regarding the for- mer, the material removal is one of the least-utilized fields of applica- tion within industrial robotics [3]. Waterjet and laser-cutting are com- monly robotically controlled, as well as robotic milling of softer materi- als (wood, urethane, sand-stone, aluminium or cast iron), but not hard materials. Insufficient robot stiffness (approximately up to 100 times less than CNC machines) is commonly identified in the industry as a limiting factor for robot machining of hard materials.

The European project Hephestos, with its focus on developing so- phisticated methods in robotic manufacturing, has produced a cost- efficient solution in hard-materials machining for small-batch produc- tion of highly customised products through the application of industrial robots. The two main limitations explained previously have been ad-

Fig. 1.   Shared machining applications and robot-technology  positioning.

dressed in Hephestos providing applicable results and a relevant step- change towards industrial applications.

Hephestos has developed a paradigm that shall provide standard industrial robots with break-through techniques in production plan- ning, programming and real-time control systems. Based on established computer-aided-manufacturing frameworks, Hephestos has optimised production planning through the automatic generation of robotic pro- grammes, while considering specific robot signatures, i.e. robot system kinematic and dynamic characteristics, as well as models of processes (e.g. milling, grinding, and polishing), that are essential for the robotic application in hard-material machining. To accommodate small-batch production time scales, real-system data obtained by means of advanced sensor techniques have also been integrated in the planning to improve efficiency.