CERN, the European Organization for Nuclear Research and the world’s largest particle physics laboratory located in the Franco–Swiss border near Geneva is most famous for employing the inventor of the World Wide Web, Tim Berners-Lee. Coined the “Web”, the intent was to create a platform for information sharing between scientists working in different universities and institutes all over the world. Once CERN had created this platform and the various scientists began collaborating across the web, the next evolutionary step was the automation of business processes.
In the early 1990’s, CERN understood how critical business process automation was to the success of their mission. With employees and contractors coming in and out of the organization frequently as new projects started and existing projects completed, CERN needed to make business processes in the area of human resources, logistics, and procurement easier and more efficient so the workers could spend more time solving problems and less time on administrative tasks.
What CERN came up with was their EDH application; providing their employees and contractors with a one stop shop portal for all of their administration needs, integrating over 50 business processes ranging from vacation requests, purchase orders, employee onboarding, rights management, to safety procedures.
CERN’s early business process automation initiatives consisted of custom development written internally using C. As time passed, more one off proprietary solutions were built which led to increased maintenance costs and less flexibility. In 1998, CERN began leveraging Oracle’s work flow engine to reduce the amount of custom code and begin leveraging business process modeling. This served its purpose for a number of years but by 2006 the existing Oracle tool was at end of life and although it met the needs from 1999 to 2006, it did not meet the needs going forward, particularly with the emerging desire to provide better integration between systems.
In 2006, CERN embarked on an effort to replace the workflow engine, and simultaneously adopt a more service-oriented approach. Derek Mathieson, the leader of the business unit behind this initiative, took some time to discuss the project with me. Derek stated that his team had over 50 business processes that needed to be updated with the intent of automating as much as possible and focusing on adding value, not administrative overhead. At the same time they wanted to provide as much self service capabilities as possible so the users could quickly handle their administrative tasks without being highly dependent on other people and processes.
So in 2006, CERN did a full vendor evaluation of BPMS tools. CERN embarked on thorough vendor evaluation process that included eight initial vendors. Eventually the list was narrowed to three vendors that matched their requirements. At that point, CERN brought in each vendor tool to perform a proof of concept and eventually selected Active Endpoints’ ActiveVOS as the tool most closely matching their needs. Some of the more important requirements were process versioning, xpath design tools, restart capabilities, and halt on fault functionality. One of the key differentiators according to Derek was ActiveVOS’s ability to detect faults in the business processes and then allow a mechanism for recovering from them. CERN must account for numerous exceptions in the normal flow of their business processes due to the large number of external entities that they interact with. With ActiveVOS, they were able to halt a process that encountered an exception, address the issue, and continue on with the remaining process flow. With the other tools that they evaluated, these exceptions would have forced them to start the entire process flow from the very beginning. This key capability is a huge time saver and one of the major deciding factors in the vendor evaluation process.
Since CERN had been automating business processes for over two decades, they did not encounter much resistance to the new process automation initiatives. In fact, the users actually expected or even demanded that the processes were automated and easy to use. The biggest challenge was on the technical side. In order to provide a consolidated single point of entry for all of these administrative processes, data had to be integrated from many different systems. Derek’s key advice for companies embarking in similar initiatives is to “get the corporate data right” first and then build the business processes on top of the data. He added that if you “don’t get the data right”, people won’t trust the system. This is a key take away that is worth repeating. You can buy the best BPMS tools, have the best project management and IT governance in place, and solve all of the cultural and resistance to change issues, but if the underlying data is not accurate none of that will matter. First impressions stay around for a while and if the users’ initial experience is inaccurate data, user adoption will suffer immensely.
With the help of the ActiveVOS BPMS solution, Derek and his team successfully achieved their goal of creating an integrated, self service portal containing automated processes for handling numerous administrative tasks and freeing up valuable time for their employees and contractors to focus on their important research. When asked about some of the biggest challenges, Derek mentioned the communications between IT and the business, specifically as it pertains to BPEL (business process execution language). According to Derek, BPEL is a low level language that is necessary for IT to effectively build and expose business processes as services. The downside is that the business doesn’t want the low level discussions and the IT to business communication becomes challenging because of the detail required. As Derek put it, the details do not “mean anything” to the business.
With the successful implementation of these business process automation and integration projects, Derek’s big challenge now is keeping up with user requests. The business has seen the value of this initiative and understands how quickly IT can react to changes in the business. The good news is that Derek’s team can have working prototypes up in days or weeks now as opposed to months and can deliver solutions much quicker than before. The challenge is that his organization is the single point of call for all business processes and keeping up with demand is an uphill battle. In this day and age of financial turmoil, I think that is a good thing. Companies investing in increasing efficiencies and lowering costs are the ones that will withstand the current financial pressures and will be that much further ahead in the future.