The S of IoT networks
The assumption about big data analytics is that if you capture more data and analyze, the more accurate your results will be. But it doesn’t matter how big your data volume is and how optimized and efficient your analytics algorithms are, if the analyzed data is inaccurate from the sources, then it will surely result into inaccurate analytics.
Sensors are the eyes and ears of the entire IoT ecosystem. Sensors convert a wide range of physical parameters into invaluable data required to run businesses efficiently. Sensors combined with on-chip processing unit and communication interfaces are called ‘smart’ sensors. Most of the signal conditioning is being handled by on-chip components. The data collected through the sensors are passed to a scalable data storage and an analytics engine for further analysis to gain invaluable insights.
The prevalent high speed and low power microprocessor technology has enabled smart sensors to manage many self-managed tasks on-chip. The addition of on-chip analytics engine which can analyze the captured data locally forwards it ahead only if the data is accurate. This will reduce the volume of data being generated by a significant margin resulting into lesser load on subsequent data communication networks. Evaluating the quality of analytical output is essential if the output is to be utilized in practice. The distributed analytical approach of putting analytics at both ends – source as well as destination, will greatly enhance the reliability of the output. With this approach, the sensor will not only introduce itself but will also inform if it doesn’t feel healthy. In future, when billions of sensors are going to be deployed in IoT networks, it is imperative that the sensors are self-aware otherwise finding the bad sensors from the good ones will become a difficult task.
Internet of Things (IoT) is defined as an interconnection of objects embedded with sensing, computing, communication and limited analytical capabilities. The four essential capabilities are the four pillars of Internet of Things technology.
Bad data leads to bad decisions
Sensors are at the core of end-point IoT devices or sensor nodes. Sensors are transducers which convert physical, non- electrical signals to electrical signals to feed our computers. The success of IoT is largely dependent on accurate measurement of physical parameters. Decisions taken by managers to run their business and consumers to control their appliances are going to be inaccurate, if it is based on inaccurate measurement of physical parameters.
Selection of the right sensors to acquire data at the first hand is a tough call. A data centric view of IoT will focus more on the selection of right sensors for right measurements.
Power is everything
Computers have transformed from big boxes to tiny chips in the last few decades. These single chip computers are called microcontrollers. The integration technologies have made it feasible to pack more power into smaller chips. MCUs running at GHz speed and ultra-low power levels is a reality now. System on Chip (SoC) centric design approach, where memory, bus interfaces and analog components are also integrated on the same chip has enabled designers to design compact and powerful computing systems running on small-sized batteries. The balance between performance, power consumption and cost will lead to energy efficient design and extended battery life.
However, the real power is not the computer but the battery. End-devices, also called sensor nodes, are responsible for collecting environmental data. These billions of IoT nodes are going to run on batteries. An intelligent system with no power to run is nothing more than a piece of junk. As we know, the computing part of a system consumes most of the power, so the selection of right MCU platform is the key to transform any IoT dream into a reality.
Chinese whispers is a game in which one person whispers a message to the next person, which is then passed through a line of people until the last player announces the message to the entire group. Inaccuracies start accumulating in the retellings, so the statement announced by the last player differs significantly, and often amusingly, from the one uttered by the first.
The ubiquitous presence of wireless communication infrastructure and the availability of energy-efficient communication modules coupled with ease of integration with MCUs is creating a storm of connected things. The longer the data travels more is the chance of data getting corrupted. Lack of standardization, data security and quality of service issues are major concerns. A careful selection of right protocols and network architecture is going to save us from making IoT a Chinese whispers game.
Intelligence – in Things or in Cloud?
IoT has the potential to transform our lives and also run our businesses more efficiently. With billions of sensor nodes generating massive amount of data, collection and storage is a big challenge. This unstructured and heterogeneous set of data has been termed as Big Data. Gaining valuable insights from big data to take real time decisions is like finding a needle in haystack. The explosion of big data is forcing organizations to upgrade current tools and processes.
Why compare human body with IoT networks?
From the perspective of an engineer, a human body is a large system consisting of many smaller systems. The cardiovascular system circulates blood around the body delivering oxygen and nutrients to organs and cells. The muscular system enables the body to move using muscles.
Humans mainly possess five senses – sight, hearing, taste, smell, and touch. The sensory organs acquire information from the environment and then sends it to the brain. The brain analyzes this information and makes decisions to generate control signals to make the body work. The peripheral nervous system connects to every sensor in the human body. The signals from sensors hop through nerves to reach the central nervous system. The arrangement of peripheral nervous system is just like mesh networks. The central nervous system is information highway made up of a bundle of nerves taking the information to brain. Brain is the most complex organ in the human body. The information collected through sensors gets stored in various parts of the brain depending on its usages. Brain’s analysis results into plan of actions that are sent to muscles to produce mechanical actions.
Similarly, IoT is defined as a network of smart objects. The sensors nodes acquire environmental information such as temperature, pressure, sound and images. This information passes through the networks hoping from node to node (a typical mesh network). Finally, this information is sent to a cloud using a high-speed internet connection as information highway. The cloud is the powerhouse of an IoT network. A cloud provides storage and analytical services in the IoT network. Offloading the main computing work to cloud (brain in human body) and deploying sensing and actuation at the edges is one of the main strategies being used in IoT networks. The cloud plays the role of central computing system in IoT networks. It stores and analyzes the information to derive result. The human body is a collection of very efficient systems working in synchronization to perform various tasks. The analogy will help us in formulating strategies and developing technologies to make IoT networks as efficient as a human body.