Integrate existing systems at lower cost

Manufacturers with disparate legacy systems running different departments, or even ERP systems that have never been fully rolled out should be looking at so-called hub technology to get their systems communicating and providing the integration they need. So says Yves de Montscheuil, director of product marketing at integration, migration and data warehouse developer Sunopsis. “They want more agility with suppliers, inventory, production or whatever, so they need to exchange data in real time. They could go for new a ERP system but that’s expensive and it rarely solves all the problems and doesn’t always work. “Going the hub route, companies typically integrate what they have and then replace when they need to application by application. That way they can do what they need to one at a time without huge expense and without closing the factory for six months.” de Montscheuil concedes that the EAI (enterprise application integration) market is very fragmented in terms of technologies, but says that’s precisely why Sunopsis has brought them together under a single hub-based platform. “We have a common user model based around a standard set of connectors – all service-based and taking into account the business rules,” he says. “You just define what you want to do and it takes care of how to do it.” In fact, the hub is the core here: de Montscheuil describes it as real-time. “Whenever an event is detected, like a new order, it’s propagated to the hub. The hub decides who is a subscriber to the information, and sends it to shipping etc at whatever rate and frequency is appropriate. “It also works with a relational database so it can be accessed by any other system connected and with permissions. You can also put a service layer on top so it acts as a single set of consistent services.” One useful word of warning though. “When you integrate systems which change data and events it’s critical to ensure that quality of data is good – poor data propagating between systems is big bad news. So for us, data quality rules are applied on the fly – like a data quality firewall – and bad data is excluded.”