Manufacturers Unify Networking Hardware and Reduce Costs

Al Presher, Contributing Editor -- Design News, June 15, 2009

What is the key trend in automation/motion networking?
Ethernet is still the hot topic in automation and motion networking, as manufacturers see a way to unify networking hardware and reduce costs. Now, we're also seeing some unification of the underlying protocols and how they are implemented at the transport layer. If we can achieve a common format at that layer, it becomes easier to support more than one protocol over a given network. But there are still challenges, especially with motion control on the network, because of the real-time needs of
the motion subsystem.

What are possible solutions for achieving real-time motion on Ethernet?
One approach is to use standard hardware, but utilize non-standard data handling that optimizes speed and synchronization. Another solution is using standard, unmodified Ethernet and encouraging, as much as possible, the use of Internet and transport protocols such as TCP/IP and UDP.  The goal is to provide interoperability and reasonable infrastructure costs, while also maintaining the performance of the motion network. What we have tried to do in the SERCOS organization and community is to figure out the best of both worlds. We are trying to provide engineers with a network that provides interoperability with other networks, without a major, negative impact in cost and ability to interface a wide range of devices.

What is the most important development goal?
The primary goal is a single-use bus for motion, I/O and factory communications. One important issue with motion networking is development of the application layer. Getting intelligent devices to talk to each other over the network is not easy, which points to the importance of profiles. No one company can provide every type of motion control or I/O device. Users expect this cross-vendor interoperability will always be true and are surprised when it doesn't work that way. There can easily be sub-features in a device that are not unique to the vendor, but since it hasn't been standardized, the feature is implemented differently in different products. If suppliers can implement features inside the device as they see fit, but use a common language to communicate with the outside world, we can go far with that concept.  

What are new technology developments with SERCOS III?
SERCOS recently released an open-source master software driver library. This simplifies developing a SERCOS-compliant product, while leveraging the development community's combined expertise in improving the library. Much has been done on defining profiles, as well. We always recognized the importance of the application layer, but didn't group idents into logical subsections and profiles that apply to different industries and types of products. With SERCOS III, there was interest from the I/O community to refine the profile for I/O. There is also a desire to attach feedback devices directly to the network rather than interfacing the feedback to the network through a drive. A current goal is a device profile and subset of idents on how to use and synchronize feedback devices.

What are areas where automation networking is improving?
SERCOS III's hot-plug functionality is a technology that is very interesting and hasn't been exploited effectively in the field yet because the standard has just been released early this year. The redundancy feature offered is another important area. By setting up SERCOS with a counter-rotating telegram, it's easy to detect a break in the network or if a device has failed. Since we were already sending telegrams in both directions around a ring and receiving them from both directions, the system can recover in a time period shorter than the shortest telegram.