What is Cognitive Computing?In a general sense, the term cognitive computing refers to new hardware and/or software that mimics the functioning of the human brain and helps to improve human the decision-making process. Cognitive computing in a broader sense describes technology platforms that are based on the scientific disciplines of artificial intelligence, signal and image processing. These platforms encompass machine learning, reasoning, natural language processing, speech recognition and vision (object recognition), human-computer interaction, among other technologies to perform specific, human-like tasks in an intelligent way.
Cognitive computing makes a totally new class of problems computable. It addresses complex situations that are characterized by ambiguity and uncertainty to handle human kinds of problems. In these dynamic, information-rich, and shifting situations, data tends to change frequently, and it is often conflicting. The goals of users evolve as they learn more and redefine their objectives. To respond to the fluid nature of users’ understanding of their problems, the cognitive computing system offers a synthesis not just of information sources but of influences, contexts, and insights. To do this, systems often need to weigh conflicting evidence and suggest an answer that is “best” rather than “right”.
Cognitive computing systems make context computable. They identify and extract context features such as the hour, location, task, history or profile to present an information set that is appropriate for an individual or for a dependent application engaged in a specific process at a specific time and place. They provide machine-aided serendipity by wading through massive collections of diverse information to find patterns and then apply those patterns to respond to the needs of the moment.
Five key technologies used to implement Cognitive Automation
- Robotic process automation: “A software automation tool that automates routine tasks such as data extraction and cleaning through existing user interfaces. The robot has a user ID just like a person and can perform rules-based tasks such as accessing email and systems, performing calculations, creating documents and reports, and checking files.”
- Smart workflow: “A process-management software tool that integrates tasks performed by groups of humans and machines (for instance, by sitting on top of RPA to help manage the process). This allows users to initiate and track the status of an end-to-end process in real time; the software will manage handoffs between different groups, including between robots and human users, and provide statistical data on bottlenecks.”
- Machine learning/advanced analytics:“Algorithms that identify patterns in structured data, such as daily performance data, through ‘supervised’ and ‘unsupervised’ learning. Supervised algorithms learn from structured data sets of inputs and outputs before beginning to make predictions based on new inputs on their own. Unsupervised algorithms observe structured data and begin to provide insights on recognized patterns.”
- Natural-language generation (NLG): “Software engines that create seamless interactions between humans and technology by following rules to translate observations from data into prose. … Structured performance data can be piped into a natural-language engine to write internal and external management reports automatically.”
- Cognitive agents: “Technologies that combine machine learning and natural-language generation to build a completely virtual workforce (or ‘agent’) that is capable of executing tasks, communicating, learning from data sets, and even making decisions based on ’emotion detection.’ Cognitive agents can be used to support employees and customers over the phone or via chat, such as in employee service centers.”
BackgroundThe definition of cognitive computing was developed in mid-2014 by a consortium of experts from BA-Insight, Babson College, Basis Technology, Cognitive Scale, CustomerMatrix, Decision Resources, Ektron, Google, HP Autonomy, IBM, Microsoft/Bing, Next Era Research, Oracle, Pivotal, SAS. Saxena Foundation, Synthexis, and Textwise/IP.com. This project was led by Sue Feldman at Synthexis and Hadley Reynolds of NextEra Research. It was sponsored by CustomerMatrix, HP Autonomy, and IBM. The goal of the project was to define how cognitive computing differs from traditional computing and to provide a non-proprietary definition of cognitive computing that could be used as a benchmark by the IT industry, researchers, the media, technology users, and buyers.
FeaturesIn order to achieve the human level of computing, cognitive systems must have the following features:
- Adaptive:They must learn as information changes, and as goals and requirements evolve. They must resolve ambiguity and tolerate unpredictability. They must be engineered to feed on dynamic data in real time, or near real time.
- Interactive:They must interact easily with users so that those users can define their needs comfortably. They may also interact with other processors, devices, and Cloud services, as well as with people.
- Iterative and stateful: They must aid in defining a problem by asking questions or finding additional source input if a problem statement is ambiguous or incomplete. They must “remember” previous interactions in a process and return information that is suitable for the specific application at that point in time.
- Contextual: They must understand, identify, and extract contextual elements such as meaning, syntax, time, location, appropriate domain, regulations, user’s profile, process, task, and goal. They may draw on multiple sources of information, including both structured and unstructured digital information, as well as sensory inputs (visual, gestural, auditory, or sensor-provided).
- Information adept:capable of integrating multiple heterogeneous sources and then synthesizing ideas or answers from them.
- Dynamic and adaptive:learn and change as they receive new information, new analyses, new users, new interactions, new contexts of inquiry or activity. Probabilistic:discover relevant patterns based on context, predict the probability of valuable connections, and return answers based on learning and deep inferencing. Find unexpected patterns: a kind of machine-aided serendipity.
- Highly integrated:all modules contribute to a central learning system and are affected by new data, interactions and each other’s historical data.
- Meaning-based:leverage language structure, semantics and relationships.
Cognitive systems differ from current computing applications in that they move beyond tabulating and calculating based on preconfigured rules and programs. Although they are capable of basic computing, they can also infer and even reason based on broad objectives. Beyond these principles, cognitive computing systems can be extended to include additional tools and technologies. They may integrate or leverage existing information systems and add a domain or task-specific interfaces and tools as required. Many of today’s applications (e.g., search, e-commerce, eDiscovery) exhibit some of these features, but it is rare to find all of them fully integrated and interactive. Cognitive systems will coexist with legacy systems into the indefinite future. Many cognitive systems will build upon today’s IT resources. But the ambition and reach of cognitive computing are fundamentally different. Leaving the model of computer-as-appliance behind, it seeks to bring computing into a closer, fundamental partnership in human endeavors.
Potential BenefitsFrom improved flexibility to higher employee morale can extend the value of cognitive automation:
- Decreased cycle times and improved throughput: Software robots are designed to perform tasks faster than a person can and do not require sleep, making 24x7 operations possible.
- Flexibility and scalability: Once a process has been defined as a series of instructions that a software robot can execute, it can be scheduled for a particular time, and as many robots as required can be quickly deployed to perform it.
- Improved accuracy:Robots are programmed to follow rules and do not make typos.
- Improved employee morale: The tasks and processes most suitable for automation are typically the most onerous and least enjoyed by employees. Employees relieved of these tasks can focus on more important and rewarding work.
- Detailed data capture: The tasks performed by machines can be monitored and recorded at every step, producing valuable data and an audit trail that can support further process improvement and help with regulatory compliance.
Cognitive Process Automation vs. Robotic Process AutomationMost people are familiar with process automation. It’s been around since Henry Ford invented the assembly line. So what is RPA? Analysts write, “The most important thing to say about RPA is that it is not a robot! At least it is not a physical robot. RPA is a type of software that is able to interface with computer systems in the same way as a person does. RPA software is able to ‘type’ and is able to ‘click’ and is able to move a cursor. This enables it to open and close programs and to use programs. This is why the term ‘robotic’ was coined — there’s no physical robot but the software behaves in a robotic way. What is key though is that the RPA software is able to carry out tasks with a much greater level of efficiency than a human operator — and it never gets tired. RPA has been shown to be a highly effective option for carrying out certain types of tasks. It has delivered huge cost savings for organizations and eye wateringly massive returns on investment of 100s of a percent in some instances. It is certainly worth every organisation taking a serious look at how they might take advantage of what it is able to do., Robotic process automation is a technology that lets software robots replicate the actions of human workers for routine tasks such as data entry and it is altering the way organizations handle many of their key business and IT processes.
RPA has been around for some time. But businesses want to move beyond RPA and make their processes smarter not just automated. They want Cognitive Process Automation™. When RPA is used in conjunction with cognitive technologies, its capabilities can be significantly expanded. The integration of cognitive technologies with RPA makes it possible to extend automation to processes that require perception or judgment. With the addition of natural language processing, chatbot technology, speech recognition, and computer vision technology, for instance, bots can extract and structure information from speech audio, text, or images and pass that structured information to the next step of the process. Analysts call this next step “intelligent process automation.” They assert intelligent process automation will be a core part of companies’ next-generation operating models. Whatever you want to call it, cognitive technologies are going to make process automation activities smarter.
ConclusionsCognitive computing systems must be able to handle ambiguity and a shifting set of variables, constantly re-evaluating information based on changes in the user, task, context, goal or new information. They must understand the question or context before seeking answers. They may offer multiple “good” answers that are weighted for confidence or closeness to the query or topic. Users must be able to interact with the system easily in a kind of continuing “conversation.” Like humans, these systems must be dynamic, and they must learn.
Cognitive computing should redefine the relationship between people and their digital environment. Context is the new element at the heart of this next computing frontier. The first wave of computing made numbers computable. The second wave has made text and rich media computable and accessible digitally. The next wave will make context computable.
Cognitive computing gathers contextual information from multiple sources, including your current task, your past behavior, the traffic, the weather and any other information that might be pertinent. It takes it all in but then filters it using the lens of your context. Together, the user and the system act as a team to steer information gathering and decisions as the twists and turns of an information path unfold.