国际原文

Best Practices of Hosted Learning Solutions

A New Cost Model for Distance Education in Europe?

E-Learning Strategy: A Framework for Success

E-Learning Maintenance Strategies — Why You Need One

Delivering the Dream: Models for Intelligent Assistance

New Territory: Adding Video to Online Learning Offerings

Op Ed: Slowing Down to Get Results Up to Speed

Engaging Learning: Experience Basics

E-Learning Standards Update

Trend: Podcasting in Academic and Corporate Learning



Workflow-Based E-Learning: Next-Generation Enterprise Learning Technology


In the last year, a new next-generation learning technology has emerged in the enterprise market: workflow-based e-learning. Enterprise application integration (EAI), primarily Web Services, is the foundation technology that underpins this latest innovation in learning technology. Web Services is the infrastructure that has created a new concept, technology, and enterprise performance experience known as workflow. A fundamentally new type of learning technology has emerged simultaneously with that workflow.

Workflow-based e-learning emerged in a coherent form beginning in late 2002. It’s the result of an intense technology convergence vector now present in the enterprise. That convergence vector has simultaneously generated emergent workflow and workflow-based e-learning. Workflow-based e-learning is characterized by

task and work support embedded in the dynamics of the real-time workflow
real-time collaboration with people and systems
learning and performance nodes modeled with Business Process Modeling tools
short, granular bursts of learning and performance support embedded at specific nodes of a business process
dynamic generation of on-the-fly tasks as work evolves
continuous performance improvement and automated performance measurement
personalized delivery, management, and routing of tasks and task support.

Quintic is one of the first companies to explicitly merge workflow with e-learning. Its flagship product, OnQ2, is sold to customers as an integrated KM-BPM-CRM-e-learning solution or as a stand-alone e-learning product. The OnQ2 product uses workflow diagrams to map workflow and learning content. It’s unique in that it doesn’t hide the workflow diagrams within the content. Workflow is displayed side-by-side as a navigation guide to workers.

Optimizing productivity, targeting lag time

The economic imperative to define and optimize precise workflows for particular workers is being driven by the need to increase productivity. In the presence of a finite work week and a shrinking workforce, worker output is the only metric that can be modified to increase productivity. Optimizing the workflow is the only way to increase productivity—and profits.

There are two ways that workflow technology optimizes productivity: increasing output by streamlining tasks and removing the lag time from business processes.

It’s well known that achieving productivity gains by streamlining processes is becoming more difficult to achieve. Smaller increments of improvement are garnered even with the most sophisticated automation technology. This trend has forced suppliers to focus on the issue of lag time. Most workflow technologies achieve small gains in actual task output but they achieve large gains by reducing lag time.

According to Ultimus, it’s believed that 90 percent of the time required to do work is lag time; only 10 percent is actual task time. Ultimus maintains that reducing lag time is the only way to significantly increase productivity. They claim that even if task time is reduced by 50 percent, it will only have an overall impact of 5 percent on the actual process time. However, if the lag time is reduced by 50 percent, the overall process time is reduced by 45 percent.

Ways to reduce lag time include sophisticated task routing, workload balancing, exception handling, and embedded task support. One of the primary ways to reduce lag time is by shrinking the training time needed to get up-to-speed on tasks.

The AWD/Knowledge Enabler is a task management modeling tool that guides a user through a series of pre-defined steps necessary to process an item of work—reducing keystrokes and training requirements. Similarly, users of Teamplate’s Integrated Development Environment (IDE) routinely cite a dramatic reduction of developer training time and costs by virtue of using the tool.

Bloom and workflow

New tools can model and measure all three of Bloom’s learning and performance domains. Indeed, a range of products can measure performance in the cognitive domain. Products from such suppliers as Knowledge Products, Knowledge Impact, Lombardi, and Ultimus can track a worker’s performance based on the paths he or she takes in well-defined decision trees. New developments in the field of situational awareness are capable of measuring the six cognitive domain levels simultaneously.

Very new products that measure performance in the affective domain are also on the market. The affective domain deals with what a person believes and feels. Analysis of the affective domain measures a person’s emotional state of mind. Until recently, this was too complicated and expensive to measure. But it’s becoming a hot issue since large studies by the Gallup Management Journal and FranklinCovey revealed that a very small percentage of employees are aligned with company goals. Equally disturbing was data showing that a large percentage of workers (17 percent to 20 percent) are actively disengaged.

FranklinCovey sells a Web-based assessment that will determine a company's execution quotient, what it calls xQ. PeopleView, which partners with THINQ, sells an assessment technology called EQSight that’s designed to measure emotional intelligence. Their data strongly suggests that personality is a key indicator of productivity and the bottom line.

Likewise, personality tests from the big consulting firms are becoming routine candidate screening tools. Not to be confused with psychometrically-unsound instruments such as Myers-Briggs, these tools are quite sophisticated and, so far, defensible in court. Several e-learning vendors, including Qwiz and Fitability, sell these assessment services.

Then, there was BAM

Business activity monitoring (BAM) is a new form of business process analysis technology that tracks business processes in real-time. BAM is essentially a human performance monitoring and automated assessment technology that observes users as they work in business applications. Business Process Management systems design and manage the business process workflow. BAM analyzes the flow for efficiency, speed, productivity, and quality, all of which are well-known aspects of classical assessment criteria.

The technology is notoriously efficient at spotting productivity bottlenecks. Combined with BPM and BI, it assesses the exact nature of the bottleneck. Combined with Presence and Instant Messaging, BAM takes performance assessment one step further and proactively alerts managers and workers when to take action to offset performance problems. Combined with workforce optimization and so-called productivity monitoring software, BAM is a very short leash, indeed. In this context, it can work as a form of enforced automated change management.

BAM products are now tracking psychomotor skills with great precision and can measure all seven levels of the psychomotor domain defined by Simpson. Products such as XStream’s Performance Analyzer can measure mouse clicks, text entry, and the time it takes to do granular tasks in a business application.

Three types of workflow-based e-learning

There are three general product categories of workflow-based e-learning: content, analytics, and personalization.

The dominant product type of workflow-based e-learning is performance intervention and remediation content embedded directly into real-time work performance. It’s a powerful form of primary and direct performance support. It also is the leading candidate for automating and capturing the apocryphal informal learning.

Workflow-based e-learning differs from classic electronic performance support systems (EPSS) because it’s iterative and evolves as work processes evolve. There’s a strong bi-directional interaction with workflow-based e-learning. It also has robust measurement, monitoring, and modification feedback loops. This goes far beyond conventional ideas of performance support.

The products from Ultimus, Lombardi, XStream, Knowledge Products, Teamplate, Knowledge Impact, Nobilis, and Hyperwave are examples of workflow-based products that enable the design, development, and delivery of workflow-based e-learning.

Automated informal learning

It’s well known that the overwhelming majority of learning occurring in the workplace is informal in the sense that it occurs outside formal training events. The statistics vary, but between 60 percent and 80 percent of learning is achieved in the dynamics of informal events. New forms of ad hoc collaboration, such as Instant Messaging and Webconferencing, function as the virtual water cooler of the enterprise. These are forms of emergent workflow.

To be sure, there has always been the dichotomy between how work is supposed to be done and how work is actually done. Systems and procedures are put in place to structure compliance, but real work often entails breaking the rules, crafting workarounds, innovating on-the-fly, and dealing with ad hoc work tasks that were never codified.

Workflow-based e-learning products can harvest evolving performance while the work is actually occurring. For example, Millennium Chemicals used OnDemand Personal Navigator when they upgraded 4000 global users to the SAP 3.1 application. The tasks were modeled by real users in a bottom-up approach. As actual users defined tasks and processes in the tool, observers quickly discovered that the workforce used a variety of short-cuts and workarounds to perform tasks more efficiently. The bottom-up approach allowed the company to capture the organizational knowledge of workers that would have been difficult to capture in traditional task analysis or packaged end-user training.

The process modeling component of the Lombardi TeamWorks product has a unique bottom-up task management feature. The workflow engine is an interactive application and workers can create new work items or modify tasks as the business evolves. They can create their own embedded process coaches and pass them into the workflow in order to help others perform new tasks. This is a powerful informal learning capture tool.

Automated performance analysis

The immediate consequence of integrated workflow is the ability to carry out sophisticated performance analysis. Because hybrid applications are integrated with all the major enterprises applications and their databases, they can link the literal performance of a worker with enterprise data, inward-looking business intelligence, and productivity metrics.

Analytics in the form of automated performance analysis is another type of workflow-based e-learning. For the first time in the history of technology and performance analysis, the performance of a worker can be cross-analyzed and cross-corroborated with all business performance metrics. The consequence of this outgrowth is that learning and performance have become the latest key performance indicators (KPIs) in the enterprise, reaching the status of core business processes.

Three primary forms of performance are modeled, simulated, managed, measured, and modified in the workflow-enabled enterprise:

discrete work task performance (workflow)
organizational work performance (workforce)
enterprise ecosystem work performance (workspace).

According to Staffware, more than 95 percent of all business processes in the enterprise workflow require the intervention of a human. That means that task analysis in business process management (BPM) and business activity management (BAM) is overwhelmingly focused on human behavior.

Task performance and workflow analytics. It’s now possible to automate tracking of individual performance in real time. Workers are using highly personalized interfaces that map directly to the exact tasks that they’re expected to perform. Those interfaces are analyzed and monitored by technology that displays real-time performance support to workers and very specific real-time performance evaluations to managers.

Workflow analytics is identical with task analysis as understood by classic instructional design. It differs from classic task analysis in that it occurs in real time. It can analyze discrete tasks being performed by particular people at particular times and in specific places in the workflow.

There are many new products designed to analyze the workflow from such e-learning companies as XStream, Knowledge Products, and Knowledge Impact and by BPM suppliers, including Teamplate, Ultimus, and Lombardi. In addition, BAM suppliers, such as firstRain, Systar, Black Pearl, and AWD, sell robust performance monitoring technology.

Autonomy, AskMe, and Tacit Knowledge also have sophisticated workflow analytics products on the market. Each has developed expertise mining and developing technology that automatically trolls through various forms of collaboration in the daily workflow of workers to discover performance patterns. They can pinpoint expert performance with great precision. On the flip side, these tools can isolate workers that are out of the loop.

Workforce analytics measure the combined performance of humans and systems is called workforce analytics. It’s a concept that entails people and systems interacting in a non-linear cybernetic process. The target of analysis is different for workforce analysis and workflow analysis; however, the difference is rapidly blurring with the release of sophisticated convergent analytical tools.

The workforce is the collective intellectual capital residing in enterprise assets, and organizational knowledge is the aggregation of workforce performance metrics. The organization chart is really a high-level map of the roles, rather than individuals, required to run the business. As espoused by workflow pioneers such as Geary Rummler, the roles and the processes in between the organization chart are the workforce. Although, workforce is an abstraction that maps to real people, it also is a model of resources required to perform business functions. All workforce readiness and workforce management technologies use the term workforce in this sense.

The concept and the technology known as workforce analytics is gaining ground in the enterprise. New products from Docent, Saba, PeopleView, Hyperwave, Indeliq, KnowledgePlanet, mGen, Siebel, SAP, and PeopleSoft focus on measuring the impact of learning on workforce productivity. That impact is measured in terms of bottom-line increase in sales, customer retention, and revenues. It is also a pronounced form of performance assessment that can pinpoint individual and organization performance gaps.

The only way to accomplish enterprise workforce performance analysis is to have access to all the data stored in the spectrum of enterprise applications such as ERP, SCM, and CRM. Therefore, workforce analytics is completely dependent on enterprise application integration.

Workforce analytics is now virtually identical with organizational business performance. It’s the evolution of business intelligence, specifically linking outward-looking business intelligence with inward-looking business intelligence. For example, Saba has licensed the Cognos’s PowerPlay business intelligence engine for the heart of their workforce analytics product.

Workspace analytics measure the combined effectiveness of the workflow and the workforce. A very sophisticated workspace analytics technology comes from Network Physics, which looks at the workspace from data flows generated internally and externally. The product uses what the company calls real-time forensics that map performance to economic and political indicators.

Workspace analytics are identical to the methodology called informatics, which is derived from complexity theory. Informatics is defined by practitioners as the practical application of complexity science--a combination of data mining and simulation technologies. It allows business and governments to change and optimize behavior of large-scale economic or political systems based on the analysis of massive amounts of complex data.

Perhaps the most well known technology based on informatics and complexity is the SCM technology from i2.

Personalized task management

Highly personalized workflow portals are now proliferating rapidly in the enterprise, replacing conventional browsers and stand-alone application interfaces. They’re modeled, managed, measured, and modified by workflow-based process management tools, and they generate workflow applications that aggregate specific tasks performed by specific workers at specific times.

These new interfaces, often called performance portals, are performance-based and defined by job roles. Because functionality and content are delivered to workers based on their job role, e-learning also is now targeted to specific workflows and job roles. The same tools that are being used to design Web interfaces are being used to map e-learning to specific performance nodes in the workflow.

In addition, these new interfaces mark the post-browser era, post-application-interface era, and post-application-suite era. More important, identical client interfaces are no longer used in favor of role-based composite applications that are generated for specific job roles.

The interfaces to those applications are often referred to as rich clients, portals, or dashboards—but they differ from the first-generation iterations of those interfaces. These interfaces are assembled with a new type of product known as application assembly tools, which are based on BPM and Web Services, making them highly process-centric. They mix and match feature sets for workers based on business rules and workflows.

Increasingly, dashboards are being used by new forms of project management and automated task management technology. Indeed, dashboards are becoming a very sophisticated performance technology, with learning embedded in the form of team collaboration.

The Automated Work Distributor (AWD) product from DST Systems is marketed as a next-generation task management tool. The AWD/Knowledge Enabler takes control after the AWD work management delivers an item to a worker for processing. The technology then collaborates with individual workers and initiates a series of tasks that changes based on the result of the preceding step.

Perhaps the best known example of personalization in the learning industry is the Aspen Personalized Delivery module integrated in the Click2learn Aspen Platform. The Aspen Personalized Delivery module adapts learning content to an individual based on personalized rule-based feedback and context-sensitive collaboration.

Enterprise learning will never be the same

By now, it should be obvious that traditional application training or even business process training fails to meet the training needs of a worker that uses a personalized composite interface comprised of several different functions from an array of applications. Workflow is the only constant, and now it becomes the foundation and context for enterprise learning.

Workflow-based learning rides the carrier wave of workflow—or real work. As a sub-carrier encoded in the workflow, learning occurs as a by-product of interaction with the system. Unlike classical instructional methodology, there’s no subsequent learning transfer phase.

In the context of the real-time workflow, work and learning are now simultaneous. In the real-time workflow, just-in-time learning is too late.

Published: August 1, 2003

地址:北京朝阳区建国路88号现代城A座2604 邮编:100022 
电话:010-85806056\85804383 传真:010-85804384