This report updates our 2018 research agenda for automation and integration in hybrid IT architecture. Our research for 2019 on Automation & Integration: Essential Tech for Digital Business is driven by the need for enterprises to formulate an ongoing transformative approach to intelligently automate the delivery of customer value and craft new and unique competitive advantages over rivals.
While there are many ways to create competitive advantage, we believe the core tools needed to discover and effect how value and advantage are created include next-generation digital automation platforms (DAPs), robotic process automation technology (RPA), hybrid integration platforms (HIPs), and process mining technologies (PMT) platforms.
The 451 Take
It's not difficult to conclude that digital automation platforms and hybrid integration platforms are viewed by business and IT leaders as critical enablers of digital business. Moreover, two related/emerging markets are now capturing their attention to assist with digital business development. RPA exploits emerging machine learning and artificial intelligence (ML/AI) technologies to extract information from unstructured data and automate manual tasks. Process mining technology has matured as a practical means to visualize, analyze and improve business processes and human tasks. Collectively, DAPs, HIPs, RPA and PMT technologies will be the focus of the Applied Infrastructure and DevOps 2019 Research Agenda. Our approach will be to monitor and report on how each technology interacts and potentially influences vendor strategy and product design, and contribute case studies of real-world deployments of these technologies in an effort to offer guidance to those seeking to buy, invest or build these essential technologies for digital business.
Recently, we conducted a private client research survey of global business and IT leaders asking them about their technology plans to enable digital business. Among the questions we asked were, "How important is a unified digital automation platform capable of UI design, application development, workflow and process automation, rules and policy management, deployment/runtime, and ML/AI to your digitization efforts?"
Of the 362 respondents spanning government, automotive and manufacturing IT decision makers, 70% said it was either needed or they already had such capability. More specifically – 25% said it was very important and absolutely needed; 26% said it was important and needed; and 19% said it was important and already have it. Some were indifferent or uninterested – 20% had a neutral response where they can take it or leave it; 8% believed it was unimportant and not needed; and 2% were uncertain but wanted to learn more.
We also asked how important is a unified data and integration middleware platform capable of hybrid integration that includes iPaaS, enterprise service bus (ESB), extract/transform/load (ETL), messaging (queuing, request/respond, pub/sub, streaming), API management, data preparation and data quality management to your digitization efforts?
Of the same 362 respondents, 81% said it was needed or it was already in place. Specifically – 24% said it was very important and absolutely needed; 30% said it was important and needed; and 27% said it was important and already have it. Some were indifferent or uninterested – 0% had a neutral response where they can take it or leave it; 4% believed it was unimportant and not needed; and 5% were uncertain and wanted to learn more.
Digital Automation Platforms
A DAP is a set of tools and resources structured within a uniform framework to enable developers to rapidly design, prototype, develop, deploy, manage and monitor business process applications – from simple task-related workflows to dynamic unstructured collaborative activity streams, and even highly structured cross-functional enterprise applications.
To do so, DAPs are equipped with a range of new capabilities that go beyond earlier business process management, enterprise content management and application-development predecessors. DAPs enable low-code or no-code capabilities and include visual models, prepackaged templates, and graphical design capabilities with drag-and-drop tooling to build software or integrate software and IT infrastructure.
DAPs include new resources to assist in user-interface and application design, synthesize the use of new and emerging technologies found in next-generation devices, and simplify the means for collaboration among business and IT professionals to jointly design, prototype and develop applications. They enable software to be composed rather than coded.
Once designs are complete, DAPs can generate and put into production executable code, essentially combining both development and runtime IT environments into a somewhat simplified DevOps platform. When equipped with ML/AI technologies, they can make applications 'smarter' by learning from process execution and extracting insight from data payloads.
DAP vendors covered in the 2019 Research Agenda will include, but not be limited to, Alfresco, AppBus, Appian, Bizagi, BonitaSoft, BP Logix, Bpm'online, CA Technologies, Dell Boomi, IBM, K2, Kofax, Microsoft, Nintex, OpenText, Oracle, Pegasystems, Red Hat, Salesforce, SAP, Signavio, and Software AG.
Robotic Process Automation
RPA crafts software robots (bots) that can automate repetitive human activities within business processes, and may also call upon ML/AI tools to help guide automation results. The goal is to improve operational performance by reducing errors, cost and cycle times; to improve business performance toward specific or desired outcomes; and to perform at scale. Accordingly, the capabilities within RPA development platforms can range broadly.
At the low end, RPA platforms may be composed of common automation tools that include screen scrapers, macros, scripting languages and APIs. RPA vendors capable of crafting more sophisticated bots can execute rules-based tasks, follow conditional routes, and execute instructions or sequential logic.
Higher-order RPA platforms include technologies capable of dealing with semi-structured or unstructured data, such as optical character recognition, image recognition, handwriting recognition and natural language processing (NLP), among others. Some RPA platforms use more advanced ML/AI technology, and can be trained to learn from iterative examination of data input patterns. They then make suggestions based on contextual awareness, and guide human or system interactions.
RPA vendors covered in the 2019 Research Agenda will include, but not be limited to, Accelirate, Automation Anywhere, Blue Prism, Catalytic, Contextor, Figure Eight, EdgeVerve Systems, Kofax, Kryon Systems, NICE, Pegasystems, Redwood Software, Softomotive, Thoughtonomy, UiPath and WorkFusion.
Hybrid Integration Platforms
HIPs represent the next-generation of integration platform-as-a-service (iPaaS) technology to enable data exchange and interoperability across distributed and disparate on-premises infrastructure, software, cloud services, mobile devices and things that now compose modern hybrid IT architecture.
HIPs have evolved from earlier iPaaS offerings into uniform platforms that consolidate a variety of capabilities to enable system, application, data and process integration. They are built upon the integration process and connector tooling that form the basis of iPaaS technology and, depending on the vendor, add: ESB, ETL, message-oriented middleware (queueing, request/reply, streaming, pub/sub), cloud-to-ground (links with on-premises infrastructure), cloud-to-cloud, and API management (APIM) capabilities.
Several HIP vendors also enable data quality management (DQM), master data management (MDM) and big-data integration for analytics. The need for such data management will drive HIP vendors to craft new ways to create and manage metadata, so that data can be easily discovered and consumed via data hubs and catalogs. Informatica has long had such a strategy. Others like Dell Boomi, SnapLogic and Talend are following suit.
HIP vendors covered in the 2019 Research Agenda will include, but not be limited to, Azuqua, Bedrock Data, Cloud Elements, Built.io, Dell Boomi, Fiorano, Flowgear, IBM, IFTTT, Informatica, Jitterbit, MuleSoft/Salesforce, Oracle, Red Hat, SAP, Scribe/TIBCO, SnapLogic, Software AG, Talend and WSO2.
Process Mining Technology
Mature enterprises over the years have procured, developed and implemented countless software applications. Codified in and across these applications are the processes they use to run their business. Market dynamics, customer expectations and competitive rivalry frequently force the need for process improvement, redesign and even new designs in response. This requires an accurate analysis of how processes execute and perform in reality.
The PMT used to accurately visualize and understand how current business processes execute and perform has been under development for many years. But early PMT offerings were costly to implement, difficult to use, and delivered useful but potentially incomplete results. This is now changing. Current PMT offerings
can examine system data to discover process routes and data paths, determine cycle times and perform root-cause analysis to expose variance, bottlenecks and deviations. Process owners and stakeholders can use PMT to examine process hierarchies, decisions, policies, rules and tasks to improve the quality and efficiency of the process flow and outcomes.
PMT does not discover undocumented process designs per se; rather, it discovers the performance and execution characteristics of processes using the data consumed or transacted through a process and various log files from the systems of record those processes run across. The output is consumed by analytic and ML/AI tools to expose insight into execution, performance and outcomes that can drive improved alternative process designs (using DAP, HIPs and RPA). Overall, the capabilities and use of current PMT offerings have improved to become more pragmatic for use in distributed heterogeneous IT infrastructure.
PMT vendors covered in the 2019 Research Agenda will include but not be limited to Celonis, DXC Technology, EdgeVerve, Fluxicon, Fujitsu, Icaro Tech, Kofax, Kryon, Lana Labs, Minit, ProcessGold, ProM (open source), QPR Software, Signavio, Softomotive and Software AG.
DAPs use low-code/no-code techniques to create, develop and run applications in a common environment, making them highly adaptive when change is required. Thus, they will play a key role in enterprise DevOps strategy. They are likely to be the 'go to' platforms for many digital business development projects due to their relative ease of use and rapid time to deployment.
RPA vendors are less likely to survive as stand-alone vendors. Many are likely to be subsumed into other platforms, including application development, DAP, and potentially PaaS and HIPs. Vendors with RPA platforms capable of automation based on rules, conditional routing and logical operations, which can modify behavior based on their learnings via ML/AI capabilities, stand a better chance as viable independent vendors. Such RPA vendors are likely to merge with and/or acquire DAP vendors to make a more complete holistic offering.
Traditional enterprise integration toolkits now include iPaaS as the means to assimilate cloud services into enterprise IT architecture. IPaaS offerings are evolving into HIPs that will play a pivotal role in digital business, subsuming traditional integration toolkits to enable a uniform platform that includes APIM, various types of DQM, and even (potentially) business process orchestration. HIP vendors will differentiate themselves by how well they combine a variety of ESB, ETL, MOM, cloud-to-on-premises links, cloud-to-cloud and APIM capabilities – and how well they enable DQM, MDM and big-data integration
PMT is emerging as more viable technology to discover process flow within and across enterprise IT systems. Such discovery and awareness are critical to improve business process performance and outcomes, to enable automations and to craft bots. DAP and RPA vendors will seek to partner with and potentially acquire PMT vendors to stimulate use and growth of their offerings. Less likely, but with some potential, PMT vendors may acquire RPA or DAP vendors. Few, though, are currently in a position to do so.