Intelligent Workflow Research Articles

EHive: An Artificial Intelligence workflow system for genomic analysis

BackgroundThe Ensembl project produces updates to its comparative genomics resources with each of its several releases per year. During each release cycle approximately two weeks are allocated to generate all the genomic alignments and the protein homology predictions. The number of calculations required for this task grows approximately quadratically with the number of species. We currently support 50 species in Ensembl and we expect the number to continue to grow in the future.ResultsWe present eHive, a new fault tolerant distributed processing system initially designed to support comparative genomic analysis, based on blackboard systems, network distributed autonomous agents, dataflow graphs and block-branch diagrams. In the eHive system a MySQL database serves as the central blackboard and the autonomous agent, a Perl script, queries the system and runs jobs as required. The system allows us to define dataflow and branching rules to suit all our production pipelines. We describe the implementation of three pipelines: (1) pairwise whole genome alignments, (2) multiple whole genome alignments and (3) gene trees with protein homology inference. Finally, we show the efficiency of the system in real case scenarios.ConclusionseHive allows us to produce computationally demanding results in a reliable and efficient way with minimal supervision and high throughput. Further documentation is available at: http://www.ensembl.org/info/docs/eHive/.

A high throughput metabolic stability screening workflow with automated assessment of data quality in pharmaceutical industry

One of the most commonly performed in vitro ADME assays during the lead generation and lead optimization stage of drug discovery is metabolic stability evaluation. Metabolic stability is typically assessed in liver microsomes, which contain Phase I metabolizing enzymes, mainly cytochrome P450 enzymes (CYPs). The amount of parent drug metabolized by these CYPs is determined by LC/MS/MS. The metabolic stability data are typically used to rank order compounds for in vivo evaluation. We describe a streamlined and intelligent workflow for the metabolic stability assay that permits high throughput analyses to be carried out while maintaining the standard of high quality. This is accomplished in the following ways: a novel post-incubation pooling strategy based on c Log D 3.0 values, coupled with ultra-performance liquid chromatography/tandem mass spectrometry (UPLC/MS/MS), enables sample analysis times to be reduced significantly while ensuring adequate chromatographic separation of compounds within a group, so as to reduce the likelihood of compound interference. Assay quality and fast turnaround of data reports is ensured by performing automated real-time intelligent re-analysis of discrete samples for compounds that do not pass user-definable criteria during the pooling analysis. Intelligent, user-independent data acquisition and data evaluation are accomplished via a custom visual basic program that ties together every step in the workflow, including cassette compound selection, compound incubation, compound optimization, sample analysis and re-analysis (when appropriate), data processing, data quality evaluation, and database upload. The workflow greatly reduces labor and improves data turnaround time while maintaining high data quality.

Weaving a semantic grid for multidisciplinary collaborative design

This paper describes a preliminary attempt at weaving a Semantic Grid for multidisciplinary collaborative design over the internet. An ontology-based service-oriented modelling approach is proposed for distributed management of multidisciplinary design knowledge, through which, various Grid services for distributed applications are annotated with domain resource ontology to facilitate their autonomous deployment, reuse and federation in multidisciplinary collaborative design in a meaningful and flexible way. A service-oriented multi-agent system architecture is laid out in the peer-to-peer network of a distributed, scalable, semantic service registry to address, in an open, decentralised and loosely coupled manner, the life cycle of multidisciplinary collaborative design activities involved in publishing, discovering and reusing various Grid services efficiently. An intelligent service-oriented workflow planner equipped with an ontology- and rule-based problem-solving strategy is devised to support the users in constructing the workflow of a complex engineering design process from existing multidisciplinary design resources automatically, utilising their semantic descriptions. An example in applying the proposed approach to a multidisciplinary collaborative design of metal stamping progressive die is presented to illustrate the potential application of the workflow planner in the Semantic Grid.

From Data to Knowledge to Discoveries: Artificial Intelligence and Scientific Workflows

Scientific computing has entered a new era of scale and sharing with the arrival of cyberinfrastructure facilities for computational experimentation. A key emerging concept is scientific workflows, which provide a declarative representation of complex scientific applications that can be automatically managed and executed in distributed shared resources. In the coming decades, computational experimentation will push the boundaries of current cyberinfrastructure in terms of inter-disciplinary scope and integrative models of scientific phenomena under study. This paper argues that knowledge-rich workflow environments will provide necessary capabilities for that vision by assisting scientists to validate and vet complex analysis processes and by automating important aspects of scientific exploration and discovery.

Control-path Oriented Workflow Intelligence Analyses

This paper proposes two kinds of control-path oriented workflow knowledge analysis approaches which will be applied to a workflow intelligence and quality improvement framework aiming at the high degree of the workflow traceability and rediscoverability. The framework needs two kinds of algorithms ⎯ One is for generating the total sequences of the control-paths from a workflow model, and the other is for rediscovering the runtime enactment history of each control-path out of the total sequences from the corresponding workflow’s execution logs. The proposed approaches have something to do with the former, and each of them include a set of detailed algorithms for analyzing the total sequences of the control-path perspective of a workflow model. Eventually, the analyzed results from the approaches will be effectively used for some rediscovery approaches to rediscover control-path oriented workflow intelligence from the runtime enactment history, which is called workflow events log. Based upon those control-path oriented workflow intelligences, it will be realized for a workflow model not only to be gradually refined, but also to be finally revealed to maximize its quality by repeatedly redesigning, reengineering and/or refining during its whole life-long time period.

Real-Time Field Surveillance and Well Services Management in a Large Mature Onshore Field: Case Study

Summary This paper describes the planning for, implementation of and results generated by a real-time field surveillance and well services management system, as it was deployed in an onshore mature field in California, USA. The motivation behind the deployment of this system was simultaneously to improve efficiency and reduce operating costs in this large field with over 1,000 wells. The paper will describe how the business processes and supporting workflows were defined. This is an essential step before any technology can be deployed. The challenges of data management included not only the automatic handling of very large quantities of real-time data, but also the management of inventory and the integration of field level data with corporate level data. Historical data had to be brought into, and made compatible with the new system. The technologies required for this project included the software systems and the integration of these with remoteintelligent field sensors and data transmission systems. The impact of the system has been material to the performance of the asset. Examples will be given of tangible improvements in performance across the disciplines of surveillance, production engineering, and well services. One critical factor to the successful deployment of this system includes the organizational changes needed to support the new working practices enabled by the system. The paper will discuss the required change management programs. The success of this project has clearly established that a "smart" solution integrating intelligent remote devices, communications networks and workflow management software can be successfully deployed on large, mature fields. The deployment process to achieve this has been assimilated and is now being reproduced in many other similar fields across North America. The paper will indicate some of the areas where this combination of technology and supporting change management will be expanded in the future.

A framework for supporting intelligent and flexible workflow-based engineering design process

To support an efficient management of a creative and dynamic Engineering Design Process (EDP), this paper proposes a new intelligent and flexible workflow-based EDP management framework entitled EnDesProM. The Autonomic Object (AO) based on autonomic computing is embedded in the nodes of flexible workflow to ensure the intelligence of the framework. An innovative flexible workflow modelling method based on dynamic instance creation is proposed as part of the EnDesProM. Both the formalised definition and graphical expression of the EDP models are introduced to ensure the rigour of the approach. The architecture of the intelligent workflow is designed and the application of the weighted Certainty Factor (CF) model in the knowledge expression is investigated to deal with the incomplete information in EDP. The EnDesProM has been evaluated through an EDP in a Make-to-Order (MTO) company and the results show that EnDesProM can support the creative and dynamic EDP satisfactorily, though further improvements are required.

EWD-P as an example of a "The Best of Good Practice" project

The European Document Exchange System – Poland (EWD-P), developed by Rodan Systems SA is one of the most modern electronic document exchange systems in European Union. EWD-P is a workflow system that facilitates decision-making within the state bureaucratic system, with a special brief to work out official Polish government standpoints on numerous legislative issues constantly arising within the EU. The EWD-P project has created an effective platform for electronic exchange of documents related to the EU legislative process. The intelligent workflow management functionality aims to support a complex flow of documents through the meanders of the central government administration. The EWD-P system includes a high-level classification of documents using an artificial intelligence technique for document categorization. The EWD-P system simplifies interaction between ministerial departments involved in elaboration of a final common position and facilitates a more efficient organization of work within government and other public administration institutions involved in the EU legislative process.

Developing an agent‐based workflow management system for collaborative product design

PurposeThe purpose of this research is to develop a prototype of agent‐based intelligent workflow system for product design collaboration in a distributed network environment.Design/methodology/approachThis research separates the collaborative workflow enactment mechanisms from the collaborative workflow building tools for flexible workflow management. Applying the XML/RDF (resource description framework) ontology schema, workflow logic is described in a standard representation. Lastly, a case study in collaborative system‐on‐chip (SoC) design is depicted to demonstrate the agent‐based workflow system for the design collaboration on the web.FindingsAgent technology can overcome the difficulty of interoperability in cross‐platform, distributed environment with standard RDF data schema. Control and update of workflow functions become flexible and versatile by simply modifying agent reasoning and behaviors.Research limitations/implicationsWhen business partners want to collaborate, how to integrate agents in different workflows becomes a critical issues.Practical implicationsAgent technology can facilitate design cooperation and teamwork communication in a collaborative, transparent product development environment.Originality/valueThis research establishes generalized flow logic RDF models and an agent‐based intelligent workflow management system, called AWfMS, based on the RDF schema of workflow definition. AWfMS minimizes barriers in the distributed design process and hence increases design cooperations among partners.

Shaping the Industry's Approach to Intelligent Energy

SPE's first Intelligent Energy conference, to be held 11–13 April in Amsterdam, will focus on the upstream industry's drive to improve oilfield performance through the use of digital-age technologies. More than 60 technical, three plenary, and 14 poster sessions will address the challenges of integrating these new technologies across the different workflows and areas of activity in the oil and gas industry. Program organizers chose five papers that will be presented at the conference as examples of some of the best uses of digital technology operating in the oil and gas sector today. The technical papers discussed in this article are listed at left. Accelerating Mars Hydrocarbon Development In SPE paper 99482, R. Peterson et al. state, "Real value in reservoir geoscience and engineering lies in the ability to optimize the coupling between static and dynamic components at both appraisal and development stages." Elaborating on a joint Shell-Schlumberger project to upgrade capabilities and global processes on deepwater Gulf of Mexico fields such as Mars, the authors outline how an integrated solution enables the asset team (driven by group consensus) to optimize resources on the right reservoir scenarios and the most relevant sources of risk. The project targets the area where there is the highest possibility of making decisions that could possibly erode significant value from the field—the reservoir-modeling and concept-solution phase. This area requires a high level of input from multiple asset team members. According to the authors, because an exhaustive number of uncertainties (Fig. 1) and a variety of development options must be considered, evaluation time can take years. Also, sometimes reservoir models are not developed to the appropriate level of detail to support the required decision making. Not only do the models take a long time to build, but they also identify uncertainties, and the resulting assumptions are understood only on a discipline-by-discipline basis. The authors detail how technology enablers (e.g., scenario-options evaluation, decision support systems, collaborative environments) and organizational change (intelligent workflows, R&D, and business models) are used to improve execution of hydrocarbon-development and integrated-reservoir-management (IRM) processes. A solution embodying a combination of real-time R&D coupled with strong asset team project alignment to support "right-time" decisions is being leveraged on Shell's Mars field. Solution elements include a Smart Workflow System, an Uncertainty Management Tool, and a Smart Collaborative Environment (Fig. 2) joined with new collaborative work processes. The joint project scope will take data-management issues beyond existing architecture and functionality. Also, a prototype Smart Workflow System extends integrated framework capabilities and addresses such shortfalls as nontransparent tracking, poor audit trail of the multidisciplinary decision rationale, and lack of guides and prompts/best practices from within the application suite. The prototype system employs a Web-based environment that supports human workflow and automated processes.

DEVELOPING AN INTELLIGENT WORKFLOW MANAGEMENT SYSTEM TO MANAGE PROJECT PROCESSES WITH DYNAMIC RESOURCE CONTROL

ABSTRACT This research develops an intelligent workflow management system (IWMS) to efficiently and effectively manage complex processes of projects, control project flows, and allocate available resources for project execution dynamically across organizational boundaries. Process logics are embedded in predicate (IF-THEN) rules that can be activated when condition is met to drive workflows. In this paper, an expert system inference engine is applied to intelligently reason and manage workflows as well as resource allocations in a virtual web-enabled environment. Hence, project results can be controlled precisely to achieve the project goal and to meet the required project deliverables. In this study, we go into details to describe the flow operational mechanisms for the IWMS. Finally, we use a system-on-a-chip (SOC) development project as a case to demonstrate IWMS' comprehensive functions and capabilities for collaborative project management.

SET-BASED ANALYSIS OF STRUCTURED WORKFLOW DEFINITION

An error-comprising workflow definition of mission critical business process might incur serious problems to an enterprise. Although workflow designer is responsible for the error-comprising workflow definitions, workflow system has to be equipped with an intelligent workflow modeling tool preventing workflow designers from specifying error-comprising workflow definitions. Faults and mistakes of process designers have to be detected and reported to them by the tool at workflow build time. Access conflicts and improper specification of exceptions are two typical examples of such an error-comprising workflow definition. In this paper, we develop an access conflict detection and an uncaught exception detection techniques. A simple workflow definition language, named SWDL, is developed and the techniques are successfully developed on SWDL using Set Constraint System. With slight modifications and scope restrictions, the proposed techniques can be used in any workflow definition language either by translating it into SWDL or by referring to the techniques for the developing its own techniques. This indicates that general conventional programming language analysis techniques can be used in the analysis of workflow definitions by introducing an intermediate workflow definition language and developing analysis techniques on it.

The design of intelligent workflow monitoring with agent technology

In recent years, workflow technology has been widely used in business process management. With the increased complexity, uncertainty and risks in business operations, workflow monitoring is gaining growing attention in business process controlling and supervision. However, monitoring functions provided in traditional workflow systems lack flexibility, and provide little support for managing complex and dynamical changes in business process. In this paper, we propose a novel workflow monitoring approach, in which various intelligent agents work together to perform flexible monitoring tasks in an autonomous and collaborative way. By using customized monitoring plan and proactive monitoring process, the workflow monitoring activities can be executed flexibly and efficiently. The application of intelligent agents for such flexible, adaptive and collaborative workflow monitoring is investigated through an intelligent monitoring system in securities trading.

From office automation to intelligent workflow systems

The paper traces the development of a family of task support systems. Starting with the Poise system in the early 1980s, we visited issues such as office automation and the paperless office. We showed that office information systems such as Poise suffered from procedural rigidity, resulting in limited applicability. Knowledge-based techniques then emerged to meet these limitations. The successor to Poise was Polymer, a task support system built upon a plan-based expert system. D-Polymer, the distributed version, followed, expanding the scope of Polymer to multiple users (workgroups). D-Polymer used techniques from distributed artificial intelligence for multiagent coordination. A parallel trend in workgroup support in the late 1980s and early 1990s was the tool perspective that heavily influenced CSCW systems. The tool perspective viewed systems such as D-Polymer as overly complex, favoring highly user-centered tools. Acknowledging the contributions of the tool perspective, we developed D-Polymer to address workflow issues. The most recent member in the Poise/Polymer line of research is Polyflow. This workflow system uses the power of goal-based knowledge representation to assist in planning, monitoring, and repairing workflows with emphasis on the users and their workgroups.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>