Process innovation was a central theme of last week’s IBM Impact conference – it took center stage in the Day 2 Main Tent – but the term BPM has seemingly been banished at IBM, replaced by “Smarter Process”. Now BPM is just the name of one product in the larger Smarter Process marketecture, shown below.
The Smarter Process “suite” is the white block in the middle, where BPM, together with Operational Decision Manager and Case Manager, are being extended to integrate with mobile, social, cloud, and big data, as well as with each other. That’s presumably the “smarter” part. There is also something at the bottom called Operational Intelligence, which doesn’t seem to exist yet but I think is the next incarnation of IBM Business Monitor.
BPM used to be mostly about doing work faster, more efficiently and effectively, with a bit more “business agility”. The focus was on the internal workings of the organization and improving the bottom line. Smarter process seems to be more about raising the top line, leveraging continuous engagement with the customer in the age of mobile, social, cloud, and big data. In fact, the breathless customer engagement messaging harkens back to the late 1990s and the rise of e-commerce and CRM, if you just substitute “mobile device” for “web”. I doubt that will really be the sweet spot for BPM going forward, but it has replaced SOA as the new central organizing principle of IBM’s middleware business, formerly known as Websphere.
Middleware now is about “Systems of Interaction” – the stuff that connects Systems of Engagement, like mobile devices, to Systems of Record. There’s also cloud in there and “internet of things” – hey, middleware is a broad concept. So how does Smarter Process fit in? Here is what Websphere CTO Jerry Cuomo described as the formula for systems of interaction: Detect (monitor events from the systems of engagement), Enrich (use big data to respond with useful information), Perceive (listen for the customer request), and Act (do something). BPM has always had the Perceive-Act piece. The other part is new.
It’s an interesting story, and one that IBM can tell better than anyone else. At Impact, they showed that they have many of the components already. But today, assembling them into true Smarter Processes would seem to require a substantial investment in a myriad of disconnected platforms and tools, highly skilled specialist programmers, plus a massive dose of professional services. There is something in IBM’s DNA that is always drawn to that. But that is exactly what puts it at odds with IBM’s very successful BPM story for the past three years. That Lombardi-driven revolution was based on:
- Model-driven development, not a lot of special code
- Common graphical models shared by business and IT
- Business-IT collaboration throughout the implementation cycle, featuring playback and iterative enhancement
- Business-oriented governance to stimulate growth from project to program to business transformation
Many of the customer success stories at Impact, such as Banco Espirito Santo, hinged exactly on these attributes. It’s too early to expect much of that kind of tooling ready for the larger Smarter Process story, but it was a little troubling that there was no roadmap for it at Impact, or even a statement that Smarter Process would try to preserve those values. Here is a report card for the current state of affairs.
Decision management (fka business rules) now plays a key role in IBM’s BPM story, and IBM has brought to it all those BPM values of simplified tooling, improved governance, standards-based modeling (DMN), and business-oriented “discovery” in Blueworks Live. Grade: A
Mobile enablement. A developer license for Worklight is now inside the BPM box, along with sample code, simplifying development of mobile apps and coaches integrated with BPM. In the Solution Center, BP3′s mobile toolkit looked even better, with less development effort. The Perceive and Act part of the mobile Smarter Process appears to be well on its way. Grade: A-
Insight to action. BPM 8.5 replaces the old Lombardi BAM ScoreBoards with customizable dashboard widgets based on Coach Views technology, so they can be incorporated more flexibly into coaches (task UI) and the Process Portal. I particularly like the new Gantt chart views of average completion time and instance history with projected future completion, annotated by the process activity stream. Really nice. But the full feedback loop from runtime performance to real-time remediation is mostly outside the Lombardi cocoon, involving Monitor and ODM and message broker, Integration Designer and Business Space. I would say this is the part of the platform most in need of BPM-ification. Grade: B
Integration bus. A new IBM Integration Bus unifies the functionality of ESB and Message Broker. Geeky yes, but a good first step. The way that all that Detect and Perceive business connects to Smarter Process is via ODM integrated with the bus. Eventually it would be nice to be able to model those interactions in Process Designer. Grade: B+
Donut hole. I’ve been complaining for a couple years now about IBM’s lack of modeling/analysis tools for process analysts that do more than generate business requirements on paper, i.e., the replacement for WebSphere Modeler. When I asked about this last year, the answer was “Please go away.” This year I didn’t have to ask. At one of the BPM sessions, customers – all Modeler users – were asking the same thing, and the answer was, “We’re starting to hear this; what are your requirements?” The prior BPM team at IBM did not want to ruin Blueworks Live with features for business analysts, business architects, or other professional modelers. The idea was that business users would simply create conceptual “discovery” models in Blueworks Live and then developers would use those to create physical models (i.e., implementations) in Process Designer. There was no notion of logical models – of process flows, data, decisions, KPIs, etc. – that could be exported to Process Designer. Now Blueworks Live provides multi-level decision modeling using DMN, and is talking about enhancing the BPMN tool to better match best modeling style (e.g., indicating message flows to external entities). There is even the first hint of proper model interchange between Blueworks Live and Process Designer. These are all hopeful signs. Hint to IBM: Logical models for data (entity relationship diagrams), forms, and KPIs would be a good start. Grade: Incomplete
Case management. Everyone – IBMers included, I think – expected some kind of merging of BPM and case management to be announced. It still didn’t happen; lingering warlord bickering over revenue recognition is the rumor. Unlike last year, none of the BPM analysts had the heart to ask IBM about it publicly. The IBMers seemed just as miserable about it as we were. Grade: Inexcusable
Together with Shelley Sweet of I4Process, I began doing a set of basic Process Mapping 101 training videos for Lombardi Blueprint several years back, and have updated it every couple years for IBM Blueworks Live. The latest iteration is now online, available for free whether or not you are a Blueworks Live user. In addition to the basics of how to get started in BPM – how to organize the project, staff the team, create the first high level map, and communicate with the sponsor – the latest version shows off more of the advanced capabilities of the tool, including modeling policies and decisions, and Blueworks Live’s extensive collaboration features. There’s even a bit of BPMN Method and Style thrown in. The link is here. You don’t need to be a BWL subscriber but you have to give up your name and email. Some of the boilerplate text on the registration page is still from the older version, but trust me, the content is all new. Check it out, let me know what you think.
I recently received the following from a former student in my BPMN Method and Style training:
“I’m trying to find some rules regarding when pools should be shown as a black box with message flows to the pool boundary or multiple white box pools with message flows connecting events/activities within the different pools. From what I have seen, it seems that the multi-pool method can be used when showing interaction between different processes and black box pools are used to show external process participants. Is this correct? Any clarity you can provide on this would be greatly appreciated.”
First, the context: Some business processes cannot be modeled as a single BPMN process because the instances involved in different parts do not have one-to-one correspondence. BPMN Method and Style asks you to model that scenario with multi-pool structures, each internal pool representing a single BPMN process, coordinating their actions and states through a combination of message flows and shared data stores. In the examples in class the internal pools are white-box (containing flow nodes) and the external pools are black-box (empty). The question is what is the “rule” about this.
Of course, the BPMN 2.0 spec provides no guidance. My general rule would be simply this: If the second (or third, etc.) internal pool is defined independently of its use in the multi-pool structure, it is better modeled in its own model and represented in the multi-pool diagram as a black-box pool (hyperlinked through the tool or model repository to its definition diagram). If not, it is OK to model it as a second white-box pool in the original collaboration diagram. Whether modeled as black-box or white-box, pools representing internal processes should be labeled with the name of the process, not the organization.
I recently received the following letter:
I am a Computer Science student and Software Engineer interested in scientific workflow technologies and BPMN2.0 modeling/execution semantics. If possible, I would like to ask for your kind opinion on a simple BPMN2.0 modeling issue I am wondering about. This issue is related to the following question: “is it possible to overwrite activity-task’s properties with InputDataAssociations ?”.
Suppose we model a task T with dataInput dI and property P as follows:<task id="T"> <ioSpecification> <dataInput id="dI" /> <inputSet> <dataInputRefs>dI</dataInputRefs> </inputSet> </ioSpecification> <property id="P" /> </task>
I was wondering weather or not it is both BPMN2.0 compliant and elegant to introduce the following assignment InputDataAssociation “dA” inside task “T”, in order to write P ?<task id="T"> [...] <inputDataAssociation id="dA"> <targetRef>dI</targetRef> <assignment> <from>dI</from> <to>P</to> </assignment> </inputDataAssociation> </task>
I think this modeling is not completely elegant, maybe because dI is used as a source value of the assignment, despite the fact that it is a target of the inputDataAssociation. Since the target of an inputDataAssociation can not directly be a task’s property (can be?), I wonder weather or not BPMN2.0 task’s properties can be overwritten with data values from the outside task’s scope. If you could share your kind opinion about this topic, it will be of a great help and highly appreciated. Thank you.
I am publicly on record as saying the data flow portions of the BPMN 2.0 spec are the worst part of the document, so vague as to be mostly unusable. And I have debated the student’s question previously, in a slightly different form, with Falko Menge of camunda, so even BPMN experts do not agree on the answer. The spec does not clearly distinguish between a dataObject and a property except to say that there is no graphic for a property, even though 99% of the time, the graphic for dataInput is omitted, so this is not a useful distinction. They both represent stored instance values accessible only during the lifetime of the running instance of the activity, process, or event in which they are defined. In other words, technically the spec defines only their use in executable BPMN; for non-executable, you are left to your imagination. They are both explicitly item-aware, meaning they both may be a source or target of a data association.
One difference related to the student’s question is that a data object cannot belong to a task, but a property can. The data object may belong to a subprocess or process containing the task, and its value may be mapped at runtime to the task data input by a data association. About that much there is no debate.
The debate is related to the meaning of a data input. Is it an interface specification (my view) or a stored value (Falko’s view)? The spec says on pp211-213 that data input is used to specify the input requirements of an activity (sounds like a signature or interface to me), and that a data input may have incoming data association. It does not say a data input may have outgoing data association, and if you consider it, as I do, to be an interface not a stored value, then outgoing data association makes no sense. But as I said earlier, there is no agreement on this, and given OMG’s record on such things, I don’t expect resolution of the question in my lifetime.
But I would say that neither side of that argument would favor the student’s solution of mapping a task data input to a property of the same task. The property can attain its value in one of two ways: mapping from a stored value external to the task (data object or data store), or action of the task itself. In the former case there would probably be a data input association directly to the property from the external data source; in the latter case there would be no data input association. If you consider a data input a storage location, I suppose it might be legal to populate the property via a mapping to the data input and then a second mapping from there to the property, but I don’t see the purpose in the extra step. That might make sense if the property lifetime were not limited to the lifetime of the task, for example if it could store some value until the end of its containing process instance. But it can’t.
Thanks as always to Sandy Kemsley for a detailed summary of all the talks. My review is more impressionistic. The goals of the program were these:
- Talk about something I haven’t seen before
- Actually, SHOW me don’t just tell me
- Spur my imagination
And the program in the end met those goals. It was really really good.
There were several themes in the presentations: lowering the barriers to business users; leveraging social networks and mobile devices; handling unstructured work; expanding the boundaries of process analysis.
Lowering the barriers. Signavio showed a variety of ways to create process models without drawing them, including simple tabular input, voice input, and a Tai Chi like gesture recognition. itp commerce showed automatic scoring of BPMN models for quality across several dimensions. camunda showed true roundtripping between third-party business-oriented modeling tools and a BPMS, the first I’ve seen to do that well. Trisotech showed simple but a radically different style of tooling for capturing SME input for both BPM and ACM (is there really a difference?) Knowledge Partners International showed how business users can directly model and maintain thousands of business rules using The Decision Model.
Social and mobile. BonitaSoft and WebRatio both showed integration of BPM with popular social networks, and TidalWave showed BPM applications within the Facebook environment. OpenText previewed Touch, their next-generation social/mobile platform exposed as an embeddable widget throughout their BPM/ACM portfolio. BP3 gave a great presentation on what adding mobility to BPM really means and the best ways to do it, native vs hybrid vs responsive HTML5. In a similar vein, Kofax showed how a cloud multi-tenant BPM architecture should really work.
Unstructured work. IBM surprised me with a demonstration of how adding stateful data to the process modeling environment opens up the whole world of unstructured work to the BPM platform. Now tell me again why BPM and case management are different products? Bosch showed how events from sensor-enabled “edge devices” can be processed by rules to create self-maintaining machinery. Whitestein gave a mind-expanding demo of goal-directed BPM layered on top of regular BPMN process fragments – for me this was Best in Show. Computas showed their implementation of “malleable tasks” – an award-winning solution in production – where events trigger a set of case tasks each of which can be customized at runtime by the end user. EnterpriseWeb (fka Ideate) showed an environment in which all aspects of the process are networks that can be freely interconnected by end users to manage work. It was really “out there”. And Keith Swenson gave a sneak peek at Cognoscenti, a Fujitsu skunkworks project that creates “antifragile” processes by empowering end users to freely add and interconnect projects are runtime.
Expanding the boundaries of analysis. Fluxicon showed a very polished demo of process mining, automated model discovery and analysis from event logs that ultimately won the Best in Show voting. PeopleServ showed a tool for managing the complex relationships between people in the organization, which could and should be leveraged by BPM. Lloyd Dugan and Mohamed Keshk showed how ontology tools make vast archives of process models searchable. Process Analytica showed a tool for automatically optimizing staffing allocation based on parameterized simulation. Oracle gave a terrific preview of the next-gen BAM Composer, business-oriented tooling that can generate interesting dashboards of operational data, such as trending KPI values. SAP gave an equally impressive demo of Operational Process Intelligence, able to monitor end-to-end performance spanning BPM and non-BPM systems.
I would say I learned something new from every presentation. Attendees voted for Best in Show using an app created by WebRatio. The winner was Fluxicon, followed by Whitestein, then Fujitsu. We’ll be posting the videos of all the talks to the bpmnext.com website in a few weeks, and possibly open a second round of voting or liking to the public. So stay tuned for that.
I’m just back from bpmNEXT. From my perspective as the event organizer, it could not have gone any better, and the tweets seem to agree.
- Malcolm Ross
@mrappian Great time at the #bpmnext event this week.. looking forward to next year
- Steinar Carlsen
@steinarcarlsen #bpmNEXT – Thank you @bpmswatch and @nathanielpalmer for organizing a great BPM and ACM thought leader conference – http://www.bpmnext.com
- Miguel Valdes
@miguelvaldes Back from #bpmnext. Amazing conference, speakers and discussions. Good job @bpmswatch @nathanielpalmer
- Scott Francis
@sfrancisatx @anatoly_mt @bpmswatch @nathanielpalmer had to leave to the airport – who won best in show? GREAT program – thanks for having me!
- Anatoly Belaychuk
@anatoly_mt #bpmNEXT 2013 became legend. Congrats to @bpmswatch @nathanielpalmer and to winners indeed. Thanks all for terrific company.
- Shelley Sweet
@i4process Wonderful visual analysis from data mining real data: bottlenecks, flow frequency. Thx Fluxicon at bpmNEXThttp://ow.ly/jfRkD #bpm…
- Neil Ward-Dutton
@neilwd Nice demo of alternative input modes for models from @signavio Gero Decker at #bpmnext; another tool with lots of usability thought
- Denis Gagne
@denisgagne #bpmnext Fluxicon sets the bar high with first presentation. Great uix. #bpm
- Business Process
@BPM_news bpmNEXT Opens With Paul Harmon Keynote: There’s a select group of about 80 BPM industry experts gathered toget… http://bit.ly/11gB5eI
If you follow these threads you’ll see that not only was the program great, but the F2F interaction was fantastic! We’ll do it again next year. More info to come on bpmnext.com.
We are about to launch version 6 of our BPMessentials BPMN Method and Style training across all delivery channels: live-online, live onsite, and web on-demand. The next live-online class, April 22-24, will be v6, and the web on-demand version should launch by the end of March. The new version of the training takes advantage of the new Process Modeler 6 from itp commerce, including the native model repository and the BPMN Element Ribbon in Visio. The v6 training continues the basic outline and approach of the earlier versions. In addition to learning the shapes and symbols of BPMN Level 1 and Level 2, we provide a step by step methodology for creating properly structured, easily readable, and informative BPMN 2.0 diagrams. We also present the elements of “BPMN style”, rules of model composition and element usage that make the meaning of the process logic clear from the printed diagram alone.
The new version of the training expands the discussion of BPMN’s most fundamental concepts – activity, process, and instance. BPMN has very specific meanings for those terms, which are used more broadly in other parts of BPM, but unfortunately the spec fails to explain that. This is a major source of structural problems in many BPMN diagrams, but you will learn how to avoid those problems in this class.
Another new topic is the use of event subprocesses. They are outside of the Level 2 palette, which is why we have omitted them in the past, but event subprocesses turn out to be really useful for things like global message and error handlers, non-interrupting timeouts, and other common patterns that can be modeled using boundary events, but less elegantly.
Reusable subprocesses – call activities linked to independently defined processes – get more attention in the new training, mainly because the linking mechanics are much simpler now in the tool using the built-in model repository than it was with the old filesystem linking. This is an incredibly important concept in real-world process modeling that receives almost no attention from most tools and BPMN training.
The April live-online class runs from 11am-4pm ET, or 5pm-10pm CET. The price is $1095 (1-4), $995 (5-9), or $895 (10+). It includes 60-day use of the Process Modeler for Visio tool and post-class certification. Our certification is based on an online multiple choice exam and a mail-in exercise that I personally review. The student must continue to correct it until it is perfect. Over 675 students from all over the world have been certified to date.
Open-source BPM pioneer Tom Baeyens (jBPM, Activiti) today announced the launch of his new company Effektif, aiming to dramatically lower the cost and effort of human-centric process automation. Unlike jBPM and Activiti, which targeted BPM developers, Effektif looks to empower ordinary end users with a cloud-based BPM tool. Initially Effektif will limit processes to simple flows of human tasks only with no integration. Integration with selected apps such as Salesforce and Google Apps will follow. The tool and sandbox runtime are free; deployed apps are charged some amount per activity instance, not yet announced. Effektif received initial funding from Signavio, the cloud-based BPMN tool provider (and a partner in my BPMessentials training).
This looks to me a bit like the automation piece of IBM’s Blueworks Live – the part you don’t hear that much about – but obviously Effektif has a greater incentive than IBM does to expand its capabilities into the BPM/case mainstream. Effektif allows ad hoc activities, and given his stint at Alfresco, Tom knows everything you need to know about content integration, so I would expect Effektif to wind up in the “case management” column ultimately. But they don’t want to position there right now. To start, 5-minute BPM in the cloud is the differentiation. Effektif plans initial product release this summer.
One of the things I love about Ismael Ghalimi is he is an absolute nut… in a cool, brilliant way. He launched the modern BPM era about 10 years ago (BPMN was one result of that) but his company Intalio never quite reached escape velocity. Sometime last year he started a new venture STOIC, a tool that allows end users with no more than Excel-formula skills to build mobile/web apps. There is a “process-ish” aspect to it, but it does not purport to be BPM (although I bet it could do a lot of ACM). For some reason, STOIC took their launch roadshow to a hipster hotel in West Hollywood, and – amazed that any tech company would try to plow the barren fields of SoCal – I had to check it out. I was intrigued enough to drop $50 in his Kickstarter bucket, and now I’m told that my “instance” is almost ready. Today’s missive says:
When you get your STOIC instance, you should expect nothing to work.
The great stoic philosophers knew that one of the keys to happiness was to constantly practice negative visualization. In essence, always planning for the worse to happen, then enjoying the fact that in most cases, it would not (Cf.A Guide to the Good Life).
At STOIC, we practice this ancient teaching on a regular basis. In everything we do, we try to imagine the worse case scenario, and plan around it. Today, I would like you to do the same, as you’re about to get your hands on your STOIC instance.
When I try something really new, this is my own philosophy as well, but I’ve never had the courage to ask my customers to accept it. Ten years ago, this was the state of BPM. Eventually it all works.
We still have space available in our next BPMessentials BPMN Method and Style live-online class, February 25-27 from 11am-4pm ET/5pm-10pm CET. As always, it covers the Level 1 and Level 2 shapes and symbols, a methodology for clear, consistent modeling, and elements of BPMN style. We do hands-on exercises in class, and you have 60 days use of Process Modeler for Visio to complete the post-class certification. You will learn how to do BPMN the right way! Here is the link for more information, or the direct link to register.