Internet of Things and Artificial Intelligence
Created on December 12, 2015
Approximately 25 years ago, Artificial Intelligence or (AI) as we may know it, was very visible and highly recognized in the technology world. Although Artificial Intelligence was gaining traction, it reached a point where it started losing that momentum and eventually fizzled out. Over time, Artificial Intelligence made some adaptions and found a home in such areas as Statistical Analysis and (BI) or Business Intelligence.
Today, Artificial Intelligence is recognized as the driving force behind the Semantic Web. The Semantic Web is the annotation of websites with tags which enables web crawlers to index the content more efficiently by comprehending the contents of a site. This helps companies like Google because they utilize semantic comprehension of sites to manage their contextual ad targeting and search algorithms.
Presently, Internet of Things or IoT is the new and disruptive technology in the marketplace. So far, it has had a similar path to that of Artificial Intelligence twenty-five years ago. Below is a compelling diagram generated by Accenture that was found on social media. We found it compelling because it succinctly links Internet of Things with Artificial Intelligence.
Artificial Intelligence is referenced in the title, but more importantly there are four components to the diagram that requires our attention:
- Act: A system based on the comprehension achieved in the prior step. For example, turning OFF the Sprinkler exemplifies such an action and illustrates an Internet of Things capability.
- Sense: In Internet of Things technology, there are a variety of sensors that are utilized including but not limited to heat sensors, pressure sensors, and proximity sensors. These sensors capture the pertinent information and deliver it to a cloud based system for processing and analyzing.
- Learn: Machine Learning is a component of Artificial Intelligence. For example, we reference the Sprinkler model mentioned previously in this article. In this scenario, we change the action so the Sprinkler is turned OFF, but the Weather forecast was only partially accurate and no rainfall occurred. Thus, the lawn was not watered. This learning will manifest going forward regarding this system to adjust the probabilities and weights of various rules in the Comprehension phase. Thus, future actions may learn from past actions with improved accuracy.
- Comprehend: Interpreting raw data is similar to current day BI systems that utilize techniques such as Statistical Methods, Map-Reduce, Regression, or Clustering to identify data trends. For example, data from a Sprinkler system may identify the amount of moisture in the soil. This information combined with Weather forecast data can be delivered to system that analyzes it while simultaneously analyzing that the soil needs moisture. But, because an impending storm is forecasted and the need to conserve water, the Sprinkler should be turned OFF.
The interactions of Internet of Things and Artificial Intelligence are many, and when combined, can generate powerful solutions. The NetObjex platform utilizes the four stages highlighted in the diagram above, thus offering its own disruptive and cutting edge solutions.