Thursday, July 28, 2016

IoT Will Rely Heavily on Cognitive Support

The IoT alone shows great promise for simple signal sense and respond situations. IoT evens shows more promise when combined with fixed logic, either in a chip or circuit board. Combine it with dynamic software that can late bind conditional responses and the world for IoT grows exponentially. Now let's add speed, complexity and dynamic relationships. This is when IoT will need an even bigger assist to manage it optimally. IoT will generate enough big data to sort through that it will become mind and software boggling. Enter Cognitive support to extend the reach and influence of IoT. Let's dig into this gold mine for just a few nuggets.

IoT and Speed that Kills:

IoT works at speeds that are in the nano-second range thus leaving humans in the dust with response times. The obvious answer is to add firmware or software that can deal with speed. Unfortunately responses have to be governed by constraints and sometimes human intervention even if on exception, This puts limits on agility because the firmware and software can't change around changing contexts at warp speed. People can't deal with this either though they have the rationale to understand new contexts, constraints and trends.

IoT and Machine /Deep Learning:

Machine learning can help IoT by taking the new learning situations mined out of the big data generated by IoT and emerging situations to expand conditions and responses emerging from the real world. This takes time to teach and supervise deep learning technology to be effective. While learning and supervision intense in the beginning, incremental learning can be added quickly to leverage IoT in context.

IoT and Cognitive Computing:

With cognitive computing learning can be a great assist for the IoT in that cognitive can adapt at the speed of the IoT. Cognitive can understand the nuances of human language, process questions or patterns similar to the way people think and quickly mine vast amounts of data generated by the IoT. Most importantly, cognitive computing can learn from each interaction on it's own by-passing intense training by humans.

Net; Net:

IoT and Cognitive need each other in a good way. Combine cognitive computing and compound analytics with IoT and the reach of digital becomes mind boggling.

Additional Reading: