An eye on safety

Instead of setting up light curtains to boost safety around potentially hazardous robots and production machinery, a 3D camera system can quickly achieve the same goal – and at significantly lower cost. The system can be quickly programmed to accommodate changes in layout, and is able to distinguish between somebody going into a dangerous area and the departure of an operator who has just been engaged in setting up. In the event of a shutdown, it can also identify who was responsible. The SafetyEye system from Pilz is based on units fitted with three greyscale cameras, mounted above the production areas under scrutiny. The system is being trialled on the Mercedes Benz C Class production line just outside Stuttgart, and is the result of a collaboration between Pilz and Daimler Chrysler. Pilz took responsibility for system development and provided the expertise behind the safety functionality, industrial design and programming software, while Daimler Chrysler specified the practical requirements, developed the image processing algorithms and supported the test programme. Frames from the cameras, at a rate of 20 per second, are analysed by two separate image processing systems running on the analysis unit’s two computers, each of which uses a different operating system. Currently, the unit has an Ethernet connection for communication and configuration, and a CompactFlash card slot for data exchange and backup. In future, the unit will have a SafetyNet port for the exchange of safety-related data with other safety devices or systems. “New systems based on the technology can be up and running in two to three hours and achieve savings of seventy per cent over conventional safety guarding technologies,” says Pilz product development engineer Ian Murgatroyd. Three types of zones can be set up around each machine: a warning zone; an inner zone in which entry leads to the machines being stopped; and cut zones, which exclude monitoring or fixed objects such as instrumentation cabinets. Maximum coverage is for an envelope approximating to a pyramid, with a base measuring 12.8 x 9.6m and 10m height. Up to a height of 5m, the system can detect a person’s body within the zone. If the system is only mounted 5m above floor level, giving a monitored area of 6.4 x 4.8m, the system can detect a person’s leg. For detection heights between 1.5 and 2.8m, a person’s arm can be detected. One of the strengths of the system is that it is programmable, especially as there is an increasing trend to integrate all data, nowhere more so than within the automotive industry. This could mean that design data from an original digital vehicle mock-up could be accessed to provide locations for robots to weld, and safety zones set up as part of the production line design and simulation process. Reference marks on the floor or surrounding structure are monitored by the cameras, to ensure they are correctly aligned. Lines are typically painted or taped on the floor, so operators can see the extent of warning and protection zones and avoid triggering the system unintentionally. In addition to safety monitoring, in one application that has been trialled, a robot placed components in an oven for heat treatment and then subsequently moved them to a collection bin. As well as creating a safety barrier round the robot, the analysis unit monitored the amount of product in the bin, so it could be emptied once full. There has been significant interest in using the system to improve security in museums and creating safety zones around equipment in amusement parks. Medical applications have also been proposed. Full product approval is expected to be in place this month. Subsequently, trial systems are to be offered to prospective customers. Production systems are expected to be available for September. Purchase price is £8,400 (12,500 euros) per system. Pointers * System is based on the use of three greyscale cameras in a single unit, mounted above ground and looking down on a work area * One system can typically monitor two robots * The system could, in theory, be programmed as part of an integrated factory design and simulation process