Tag: process mining

In the second half of the morning, we started with James Taylor of Decision Management Solutions showing how to use decision modeling for simpler, smarter, more agile processes. He showed what a process model looks like in the absence of externalized decisions and rules: it’s a mess of gateways and branches that basically creates a decision tree in BPMN. A cleaner solution is to externalize the decisions so that they are called as a business rules activity from the process model, but the usual challenge is that the decision logic is opaque from the viewpoint of the process modeler. James demonstrated how the DecisionsFirst modeler can be used to model decisions using the Decision Model and Notation standard, then link a read-only view of that to a process model (which he created in Signavio) so that the process modeler can see the logic behind the decision as if it were a callable subprocess. He stepped through the notation within a decision called from a loan origination process, then took us into the full DecisionsFirst modeler to add another decision to the diagram. The interesting thing about decision modeling, which is exploited in the tool, is that it is based on firmer notions of reusability of data sources, decisions and other objects than we see in process models: although reusability can definitely exist in process models, the modeling tools often don’t support it well. DecisionsFirst isn’t a rules/decision engine itself: it’s a modeling environment where decisions are assembled from the rules and decisions in other environments, including external engines, spreadsheet-based decision tables, or knowledge sources describing the decision. It also allows linking to the processes from which it is invoked, objectives and organizational context; since this is a collaborative authoring environment, it can also include comments from other designers.

François Chevresson-Aubain and Aurélien Pupier of Bonitasoft were up next to show how to build flexibility into deployed processes through a few simple but powerful features. First, adding collaboration tasks at runtime, so that a user in a pre-defined step who needs to include other users at that point can do so even if collaboration wasn’t built in at that point. Second, process model parameters can be changed (by an administrator) at runtime, which will impact all running processes based on that model: the situation demonstrated was to change an external service connector when the service call failed, then replay the tasks that failed on that service call. Both of these features are intended to address dynamic environments where the situation at runtime may be different from that at design time, and how to adjust both manual and automated tasks to accommodate those differences.

We finished the morning with Robert Shapiro of Process Analytica on improving resource utilization and productivity using his Optima workbench. Optima is a tool for a serious analyst – likely with some amount of statistical or data science background – to import a process model and runtime data, set optimization parameters (e.g., reduce resource idleness without unduly impacting cycle time), simulate the process, analyze the results, and determine how to best allocate resources in order to optimize relative to the parameters. Although a complex environment, it provides a lot of visualization of the analytics and optimization; Robert actually encourages “eyeballing” the charts and playing around with parameters to fine-tune the process, although he has a great deal more experience at that than the average user. There are a number of analytical tools that can be applied to the data, such as critical path modeling, and financial parameters to optimize revenues and costs. It can also do quite a bit of process mining based on event log inputs in XES format, including deriving a BPMN process model and data correlation based on the event logs; this type of detailed offline analysis could be applied with the data captured and visualized through an intelligent business operations dashboard for advanced process optimization.

We have one more short session after lunch, then best in show voting before bpmNEXT wraps up for another year.

The first bpmNEXT conference has finished, and it was a great experience for me. I’m still on the west coast, enjoying a half-day in San Francisco before flying home, and having a bit of time to reflect on what I liked — and the few things I didn’t like — about this week’s event.

First and foremost, this was primarily a meeting of peers to discuss BPM innovation in an open fashion. It was not targeted at customers, so there was little of the peacock-like preening and posturing that you get when vendors parade in front of them. It was also not attended by the major analyst firms, so ditto. Since the format required that presenters give a demo, there was a much heavier bias towards technical attendees, although many of them were in product management/marketing or company founder roles, so not code monkeys. For me, that’s the perfect group for networking: people who are technical and articulate, but want to talk about more than just the technology. The atmosphere was collegial and friendly, even (for the most part) between competitors, and I had the feeling that many of the presenters just wanted to show off their cool new stuff because they knew that this was the audience that would most appreciate it. I really think that Bruce and Nathaniel achieved their goal of making this “like DEMO for BPM”. For the vendors that didn’t attend, or who attended but didn’t participate because they didn’t want to show all their cool new stuff to their competitors: you are totally missing the point, and you missed a great opportunity. Ideas need a bit of exposure to the light in order to grow properly, and this is the place for that.

Second, cool and awesome demos! The “Best in Show” awards that we all voted on at the end (to Fluxicon, Whitestein and Fujitsu) were well-deserved, although many others were equally deserving. I loved the bleeding edge demos — Gero Decker of Signavio accidentally showed us how to do process modeling with head gestures — and the skunkworks projects that may never see the light of day but represent some different thinking — Keith Swenson of Fujitsu with his Cognoscenti demo really brought advanced case management to life. Anne Rozinat and Christian Gunther were in the first demo slot, which could have been tough, but they set the bar extremely high (they won first prize, after all) and wowed a lot of the North Americans there who had never had the chance to see their Disco process mining product, born out of their work at Eindhoven University. The demos that didn’t work as well were those that spent too much time on slides instead of the demo, those that were customer-facing rather than optimized for peer review, and those that tried to cover too much ground in a 20-minute demo. If I can give a word of advice to those vendors who have given me briefings in the past, treat this like that: no slides, no bullshit, no distractions.

Third, nice location, although there were some minuses as well as pluses here. Asilomar is beautiful, but we had no spare time to do more than take a quick walk to the beach on a break. I don’t propose lengthening the calendar or reducing the number of demos, but rather rearranging the first day: take a nice long break in the middle of the day after lunch for people to explore around, then go a bit later or even have some evening demos after dinner. Since pretty much everyone arrived by Tuesday dinner, there could have been demos on that evening, too. It’s a fairly remote location, people mostly didn’t do much after dinner anyway (except, for those of us from time zones further east, go to bed early), and evening demos could have been fun sessions with some libations involved — maybe a BPM buzzword drinking game? The remote location may have deterred some, it’s a 2-1/2 hour drive south of San Francisco, but it did mean that we mostly hung out with groups of other attendees, making for better networking.

Fourth, this was an unparalleled networking experience. As an extroverted introvert (really, just check my Myers-Briggs), I usually dislike networking events, but this felt less like an uncomfortable meet-and-greet and more like chatting with old friends. Which, for the most part, it was: I knew a lot of people there, both face-to-face and online. I had the chance to meet several people who I knew only online in the past but already felt like I knew, such as Anatoly Belaychuk and Ashish Bhagwat. About 1/3 of the 80 attendees were international (including three Canadians), meaning that there was a significant European contingent here, showing off some of the outstanding BPM innovation that is less often seen in North America, and possibly creating some trans-Atlantic relationships that might bloom into partnerships in the future.

Lastly, a few healthier, lower-carb snacks would not have gone amiss: I think that I ate my own weight in chocolate. Only the health benefits of the red wine offset all that. 🙂

Full bpmNEXT program here. The format is 30 minutes per speaker: 20 minutes of demo, 10 minutes of Q&A.

Day 1, first session: it’s Model Morning!

Process Mining: Discovering Process Maps from Data, Anne Rozinat and Christian Gunther, Fluxicon – uncovering the secrets within your processes

Process mining, beginning with an analogy of analyzing wind and currents to optimize sailing routes. The process reality is always more complex than the ideal model. The model is helpful for understanding the overall flow, but you need to see what’s actually happening if you want to understand – and optimize – it in more detail. Demo of their Disco product, starting with a log file in a spreadsheet that includes a case ID, start and end timestamp, activity name, resource and role for each activity in a process (resource and role are not required, but can help with the analysis). Imported this into Disco, identify the fields, and Disco reconstructs the process as it was actually executed. The colors, thickness of arcs and numbers indicate the frequency of each path – great for identifying exceptions and non-compliance. Interactive filters can remove less frequent paths to show the main flow. Statistics views show aggregate information for the data set, such as case (process instance) duration and mean wait time. Filters can be added, e.g. to identify the cases with the longest duration, then switch back to the process diagram view to see the flow for that filtered set of cases to identify the bottlenecks and loopbacks. Performance view of process diagram shows the time for each path. Animation view feeds the actual log data through the process diagram to help with visualization of bottlenecks. Clicking on a path creates a filter for only those cases that follow that path — the example shown is to find the small number of cases that bypassed a mandatory compliance step — then view the actual log data to identify the cases and resources involved. Can also generate a performance view of the log data rather than a process view, that shows the path between roles, indicating handoffs in the process and allowing for identification of resource bottlenecks. They’re working with a simulation vendor to link the models that they generate with simulation capabilities, which would allow for what-if scenarios. A number of questions about the log file: obviously, if the data’s not there, you can’t analyze it, and you might have to massage the log file prior to importing to Disco.

Managing Process Roles and Relationships, Roy Altman, PeopleServ – putting people into context

People in today’s organizations may assume multiple roles and have complex relationships with others, with can make it difficult to manage an organizational view of the company as well as people’s responsibilities within processes. Most organizations have some sort of matrix structure when it comes to projects, for example, so that a person’s manager on a project is not the same as their reporting manager. Issues such as delegation can make things more complex. PeopleServ’s Connexxions uses a rules-based approach to determine the correct role for each resource in the context of each process; it creates a functional integration between people and a technical integration between systems. Within the tool, create a context, then add a person and department and create the relationship between them (e.g., manages). Add another person and create relationship between then, (e.g., peer). Add rules, which may include links to data from other systems such as HR systems (e.g., all of the people actively working in that department), LDAP and other sources of organizational information. Contexts can be nested, so one context can be added as a node in another context. The context can be traversed starting at a node (e.g., find a person’s manager in that context, or find all people reporting to a person): this is where a lot of the power comes in, since allows identification of someone’s role and relationships in each individual context, e.g., projects, teams and departments. These contexts/traversals can be accessed/consumed from applications in order to control access and establish contextual reporting structure. People throughout the organization could create and edit contexts as appropriate: HR may manage the overall organization chart, departmental and project managers may manage roles within their teams, and anyone may manage roles for more casual groups such as social teams.

Lowering the Barriers to BPMN, Gero Decker, Signavio – BPMN made easy, and cool

Signavio is used not just for BPMN process modeling for IT projects such as ERP, CRM and BPMS implementations; in fact, that’s only 30% of their business. 40% of customers use it purely for process analysis and improvement, and the remaining 30% used it as part of their certification, risk or compliance efforts. They focus on providing capabilities for both easy creation of an initial graphical process model, then more complex tools for validation, understanding and refinement of the model, in order to lower the barriers to good process modeling. For people less familiar with graphical modeling, they can use Quick Model to enter the activities in a spreadsheet-style format, and see the diagram generated as each activity is entered or edited. The columns in the activity entry table are What, Who, How, IT Systems, Input documents, and Output documents; only the What is required (which becomes the activity name), specifying the How causes swimlanes to appear. At any point, you can switch to the full graphical editor, which allows addition of gateways and other elements, and also supports voice input. The shape palette can be filtered by the BPMN 2.0 specification, e.g., core elements only, and there is a tool to validate against Bruce Silver’s BPMN Method and Style conventions, which displays the improper usages and suggests fixes. They can also apply best practices and other style guides, such as DODAF, and can show a list of the rules and which are violated by the model. The tool can maintain and compare different versions of a model, much like a code diff either for subsequent revisions on the same model, or variants that are branched off. Models can be exported in BPMN XML and other formats, and their tool is integrated and/or OEMed into several BPM automation tools as the front-end modeler. They have an experimental feature — not yet, or possibly ever, part of the core product — using a device that detects hand movements and gestures (or, as we saw accidentally, head movements) to edit the process models; although this would not work for fine-grained modeling, it would be cool when presenting and doing some minor edits on a model in front of a larger group. They’re also working on vocabulary analysis for refactoring/correcting element labels, which would ensure that standardized language is used within the models.

Automated Assessment of BPMN 2.0 Model Quality, Stephan Fischli and Antonio Palumbo, itp commerce – a quality engineer’s dream

itp commerce is a long-standing process modeling and analysis product that uses Microsoft Visio with their add-on as the front end graphical environment, which provides a transition for existing Visio users to create BPMN 2.0 standard process models. Bruce Silver uses this in his training (he also now uses Signavio, not sure if he’ll be maintaining both), since they were the first to provide validation of models against his style guide. They also provide a number of process model quality metrics — validation, conformance, complexity, consistency and vagueness — some of which are difficult or fuzzy or both. Looking at the model quality pop-up (from their Visio add-on), it shows overall model quality with a long list of checks to turn on/off, and preset filters for process analysis, process execution, diagram interchange, and process simulation to show the quality for those intended activities for the model. Models can also be assessed for maturity level, which is an indicator of an organization’s overall process maturity; obviously, models are only one aspect of organizational maturity, but looking at a maturity model such as OMG’s BPMM can yield requirements for process models at each maturity level, e.g., KPIs would be required in models at process maturity level 4. In their Repository Explorer tool, you can run a quality report on a process chain (a set of related processes), which generates a full spreadsheet of all of the individual metrics for each process, plus aggregate statistics on quality, conformance, maturity and much more.

Awesome first session at bpmNEXT. It’s only the mid-morning break on the first day, lots more to come.