The same developmenttrend is also found in optical measurement systems carried byrobots for inspection of, for example, car bodies and car subassemblies。 For tracking sensors, it is important to optimise theuse and coordination of the look-ahead information in relationto the robot program, to be able to handle situations withlacking measurement data and to be able to perform efficientrecoveries from process failures。 In a longer perspective, the 3Dvision concepts for handling, assembly, tracking, inspection,calibration, etc。, can be further integrated into the robotcontroller and from performance point of view it could then forsome applications also be motivated to use 3D vision in therobot servo loop。Other ongoing improvements of the robot technology arerelated to the user- and application interfaces of the robotcontroller with the purpose to make robot programming,operation and maintenance simpler even though the complexityof the robot systems is continuously increasing。 Wizard-likestep-by-step concepts based on graphical representations ofrobot movements and process actions are realized also for theuse on teach pendants (Fig。 3; ABB-4, 2006), the realistic robotsimulation interface is implemented on the robot programminglevel (RRS, 2006), tools for process modelling are developed tosimplify robot programming (Skarin, Claesson, & Bergling,2004), PLC functionality for logic control and advancedprocess- and equipment control is integrated into the robotcontrollers and remote automatic data acquisition of robotproduction data for optimisation, monitoring, preventivemaintenance and fault isolation is further developed。 The faultisolation can be based on data from the sensors in the servoloops, special supervision sensors, current and voltage levels inthe drive system and virtual sensors using the observer concept。This development will certainly proceed and it is expected thatdynamic robot models run in real time will be even more usedfor fault detection, fault isolation and diagnosis, based onresidual generation and systemidentification methods (Mattone& De Luca, 2003; Ostring, 2002)。Model-based control (Sciavicco & Siciliano, 2000) has beenfound to be very important in robotics in order to fulfil thecontradictory requirements on performance improvements andcost reductions (Fig。 4)。 The ongoing development is directedtowards more complex kinematic- and dynamic models, morecomplex multiple-input multiple-output (MIMO) controlschemes, bigger variations of static and dynamic modelparameters, increasing noise- and disturbance levels, largernumber of low mechanical eigenfrequencies and enlarged non-linearities。 Even if a lot of academic research has been made onall of these aspects, a lot of applied research is needed in orderto further improve the model-based robust control of industrialrobots。 Hand in hand with the further development of model-based control there is an important development of model-based design using virtual prototyping to improve theperformance/cost ratio, reduce the development cost and tobe able to shorten the product cycles (Pettersson, 2005)。 Veryimportant is then the mechatronic design approach withdevelopment teams consisting of both mechanical design androbot control specialists。3。 Driving forces for the future development ofindustrial robot controlAs for all other industrial products, the main robotdevelopment is governed by the need of its users。 However,robots are quite unique with respect to its versatility, whichmeans that robot technology is developed towards a large spectrum of requirements from different automation conceptsand different applications (Robot Manufacturer links, 2006)。Starting with the application requirements, development isneeded to cope with requirements arising from new processestogether with requirements from improvements of traditionalprocesses。 New processes are defined here as processes that arenew in combination with industrial robots as, for example,friction stir welding, milling, high performance laser cutting,gearbox assembly, flat TV-screen handling and sheet metaldeforming。 Examples of traditional processes that ask forfurther improvement of robot performance are water jet cutting,laser cutting, laser welding, gluing, grinding, de-burring,measurement and assembly。 Sometimes robot automation getsvery popular for manufacturers working with a specificapplication or material, as an example, the plastics industryis at present very eager to use robotics。 Looking at thecustomer-driven requirements, the base requirements are ofcourse to reduce the production cost and increase the quality ofthe end products and new automation concepts are tried fromtime to time to obtain a significant increase in productivity。Examples of such concepts are compact robots, backwardbending robots, heavy load robots, multi robot control, accuraterobots for off-line programming, flexible framing robots,sensor-controlled robots and safe robots。 At the same time,there is the continuous need to improve the flexibility of the useof robots, improve reliability, improve working conditions,decrease life cycle cost, to make installation, systemintegration, programming and maintenance easier and toincrease performance (Fig。 5)。 All these requirements drivethe robot manufacturers to make a continuous incremental  development mixed with bigger projects to develop newproduct generations and new product concepts。Simultaneously with the customer- and application pullthere is of course a technology push that forces the robotmanufacturers to take advantage of the latest technologydevelopment in other much bigger product segments。 Thus, thedevelopment of both hardware and software in the PC-areamakes a big impact on the robot controller development andefforts are made to use also technology coming from thetelecommunication area。 The cost of software development andmaintenance is continuously increasing and a long software lifetime is therefore important, whichmeans that, for example, newefficient software development environments and new conceptsfor scalable system architectures, open interfaces and com-munication concepts are important drivers of the robotcontroller development。 Even if the fastest development inrobotics is found on the controller side, there is also sometechnology push on the electromechanical side。 Examples aremore efficient and bigger compact gear boxes, more costefficient motors and drive systems, cheaper carbon compositematerial and more advanced tools for mechanical design。In the long-term perspective, there will for sure be acontinued robot development for the automotive industry andits sub suppliers。What could happen here is that increasing fuelprices, pollution and other environmental problems will make itnecessary to build smaller and more efficient lightweight carsusing, for example, electric drives in combination with fuelcells。 Then the question is how this will change therequirements on the manufacturing systems including theindustrial robots。 One possibility is that spot welding of the carbody will be replaced by, for example, laser welding andsoldering (Kobe, 2001) for metal parts and riveting, gluing andfinishing of composite parts (Brosius, 2006)。 未来工业机器人控制的发展英文文献和中文翻译(4):http://www.youerw.com/fanyi/lunwen_62458.html