Cognitive Computing

The Digital Twin concept in Industry 4.0 come into practice when we addressed the question like: “Do we need to adjust the product?” and then we think “Let’s look at the original specs, at the drawings, at snapshots, at the BOM, at simulation results, at VR iterations, at the real performance parameters” – We simply say then: “Let’s look at the digital twin!”
Digital Twin is a mirroring (or twinning) of what exists in the real world and what exists in the virtual world. It contains all the informational sets of the physical ‘thing’ meaning its cross-discipline – not just a mechanical / geometric representation, but also including the electronics, wiring, software, firmware, etc. not just the CAD model.
Digital Twins are currently used in the context of monitoring, simulation and predictive maintenance which are all incredibly valuable and potentially transformative in their own right, however, there would seem to be much more to it. As products of all types move to include connectivity, sensors and intelligence we can’t just think about the data streaming back from the field.
Without accurate “Context” – Digital Twin – time series data generated during production and ongoing operation is difficult or even impossible to understand and analyze. In addition, the ability to interpret and act upon these data often require traceability to prior information from related revisions – Digital Thread. To complicate matters further as artificial intelligence / cognitive computing is introduced the necessity for the Digital Twin becomes even greater.

The aim of cognitive computing is to mimic human thought processes in a computerized model. Using self-learning cognitive algorithms that use data mining, machine learning, pattern recognition, and natural language processing, the computer can imitate the way the human brain works.
Cognitive systems analyze the huge amount of data which is created by connected devices (not just the Internet Of Things) with diagnostic, predictive and prescriptive analytics tools which observe, learn and offer insights, suggestions and even automated actions.
Cognitive Computing and Machine learning addresses the challenge of passing the boundary of traditional data analytics algorithms, which spotlights the development of swift efficient cognitive algorithms.
These cognitive or machine learning algorithms enable real-time processing of huge volume of data, deliver precise predictions of various types such as recommending right products, customer segmentation, detecting fraud and risks, customer retention etc. Cognitive Computing and Machine learning supports these functions by creating a set of cognitive or machine learning algorithms that differ from the traditional statistical techniques. The emphasis is on real-time and highly scalable predictive/cognitive models, using fully automated methods that make data scientist tasks easier.
http://www.cognub.com/