Networking
Measurement and control systems that use a single 3D sensor often have a direct interface that is 0 – 10 VDC or 4 – 20 mA, RS-485 serial or through one of the standard interfaces such as TCP/IP. In systems where there are multiple sensors, on the other hand, they can be connected through a network to handle data from these multiple devices back to a single host. Networks already exist for this including Ethernet, Profibus, CAN-bus and others.
TCP/IP over Ethernet has become an established protocol with increasing popularity due to its ability to handle higher communication speeds and its widely accepted standards. Moving from 100Mb to gigabit Ethernet (GigE) can be seemless significantly increasing the amount data transmitted and therefore the productivity of a system.
Using a TCP/IP protocol for multiple high speed sensors where the synchronization of data is critical is not as easy as just connecting multiple sensors together on the same network. The key to synchonization is contained in the actual data sent from each of the sensors back to the host. Due to the inherent nature of TCP/IP the data from any one sensor can be seperated into multiple packets to be reassembled at the host. When more that one sensor is involved, multiple packets from the mulitple sensors are all sent back to the host, which then reassembles them correctly for each sensor. The problem is in a typical system when sequential scanning takes place, the object being measured moves over the time during which multiple sensors are taking multiple images of an object that is most often moving. The more sensors in a system, the more complex this can become. Without built in synchronization, data from each sensor has no relationship to any of the other sensors. The data can be offset from sensor to sensor in time or position as the objects being measured are travelling through the sensors. For example, on a conveyor travelling at 305 mm/sec (12" / sec) using a 60 Hz sensor results in a scan every 0.017 ms where the object will move 5 mm between scans. Now multiple cameras could result in the data being offset by between 5 mm - 10 mm between sensors. This becomes more challenging as the speed of the target travels faster through the scan zone.
LMI has solved the complexity of synchronizing networked sensors with the development of the FireSync protocol. This protocol contains encoder information and internal timing information down to the microsecond. Using information shared between all devices on the FireSync network together with timing control information, all the sensors collect 3D profile information at the same instant, triggered by a clock or encoder pulse. This allows all of the data to be transmitted with timing information embedded in the packets which the host can reassemble into 3D profiles that can be truly related to each other in time and in position.
