1.1: Introduction To Physical Computing and the Internet of Science Things

Physical Computing

Physical computing is the application of computers to real world systems where computers can respond to the environment around them. There are essentially two components that are utilized in this interaction, sensors and actuators.  Sensors collect information on the environment that is processed by interactive systems that consist of microcontrollers or computers/microcomputers, that in turn activate actuators that respond to the environmental stimuli. Figure \(\PageIndex{1}\) outlines this process, and although we may not realize it, physical computing has become ubiquitous in our life and is used to control everything from automobiles to kitchen and household appliances like ovens, washing machines and refrigerators. Most sensor signals are analog in nature, like the height of mercury in a thermometer, and the data needs to be digitized before it can be processed by a computer and acted upon. Central to physical computing are electronic systems that often utilize microcontrollers integrated into Printed Circuit Boards (PCBs). In this class you will create circuits on bread boards that are controlled by Raspberry Pi microcomputers with programs that you code, and you will also solder some simple printed circuit boards that you design, code and teste with a breadboard. No prior programming experience is required for this course.

Figure \(\PageIndex{1}\): Outline of processes involved in physical computing. (Belford CC 0.0)

 

Internet of Science Things

Physical computing enables the Internet of Things (IoT) where different physical devices communicate with each other over the internet, and the Internet of Science Things (IoST)  is the application of physical computing and IoT to scientific problems and processes. Figure \(\PageIndex{2}\) outlines some of the fundamental processes and structures of an IoST system.  Note that the microcomputer/controllers on the edge of the cloud are doing physical computing and a single edge device can do both of the processes in figure \(\PageIndex{2}\), that is a single edge device can run both sensors and acutators..  The image shows two different devices just to make the processes easier to see, but both processes can be run by the same device.

Figure \(\PageIndex{2}\): IoST diagram. Note, is a process diagram and although the images shows two microcomputer/controllers, they can be the same.  In a real IoST system there will be many different microcomputer/controllers relaying different types of data from different physical locations  (Belford CC-BY)

IoST systems like air quality monitoring will deploy multiple edge devices in different [geographical] locations that can stream data to central data servers that in turn can operate on the data from all of the streaming devices.

Cloud Computing

Cloud computing often involves Machine Learning and Artificial Intelligence data analytics based on data from multiple sensors, although in the broadest sense cloud computing is data analytics (and storage) occurring in remote data centers, in contrast to a computer a user is actually connected to.  The power of cloud computing in IoST is that it can process data from multiple devices in real time and utilize computaional resources located within a remote data center 

Edge Computing

Edge computing is when an edge device performs some form of data processing.  For example, a microcontroller with an embedded camera monitoring insects might be able to perform AI calculations to identify the type of insect and stream that data, in contrast to the very data intensive image files, and when that would be considered edge computing as the data processing occurs on the edge of the cloud.  That is, the device that actually collects the data (physical computing) can also do local data processing (edge computing) and reduce the data streams bandwidth. 

 

Paradigms of Science

Microsoft Research has a free online book, the Fourth Paradigm: Data Intensive Scientific Discovery, that postulates 4 paradigms of science. Before proceeding it is prudent to look at the curriculum content of this class in the context of these paradigms of science.

• Empirical Science-1st Paradigm (1,000s of years old): The oldest experimental based science employing qualitative and quantitative analysis of the physical world around us. 
• Theoretical Science-2nd Paradigm (centuries old): The theoretical causality-based mathematically intensive reasoning that underpins our understanding of experimental data as outlined by the scientific method commonly taught in freshmen level science classes.
• Computational Science-3rd Paradigm (decades old): The extension of the first two paradigms to solve complex equations that humans alone could not compute. These may be "ab-initio" ("from the beginning") in that they are entirely based on theoretical relationships, or "semi-empirical", where they also employ empirical data. 
• e-Science-4th Paradigm (emerging): This is the newest form of science resulting from data exploration enabled through the digitization of data. This can can lead to new discoveries and predictive analytics based on correlative-reasoning without elucidating the underlying causal relationships germane to the first three paradigms. 

Physical computing is essentially applying computers to the first paradigm as systems interact with the environment, and as such this course can be an easy way to pick up programming skills. Through edge and cloud computing Physical Computing/IoST system can enable real-time fourth paradigm processes. Thus this course can provide the coding skills to function as a gateway to third and fourth paradigm scientific discovery for empirical science students.