August 21, 2018

yet-another Internet of Things post

yet-another Internet of Things post

The phrase 'Internet of Things' or IoT is overused today and becomes all-over-place jargon. I'm with you, and I'm still using it.  I have not find a better replacement, so bear with me.  If you have heard this term before, you should check your internet, even your TV provider.  IoT becomes a trendy word, and the next-hope to improve human productivity and enjoyment. And if you live in Vietnam in 2018, like myself, this term is going to be mentioned even more.

In 2015, while I were in my postdoctoral training in Arizona State University, I heard about this phrase first time.  I did not recall the instance why it was mentioned, it could be a Amazon's Dash buttons for resupply or about Echo.  2014-2016 were also the years when the personal fitness tracking devices became a must-have gadget to track step, heartbeat, sleep, and some others.  I owned Fitbit Charge and later Charge HR, so the feeling of living in the world when every equipment around me connected to internet, collected physical data,  then analyzed and displayed in an iPhone app  was reasonable and expected.   The photo below displays neatly the pathways.  In my last year of the doctoral program in the environmental biotechnology, I got to know the embedded systems, and spent time with Arduino. I developed a strong interest toward IoT devices and the application.  Writing down this topic also help me with reinforce learning, which also means that I will do research but may miss some important points here and there.


The feeling of relevance and closeness to IoT is not only the reason why I started this blog threats, and potentially created more noise than clarity on the same topic.  To move on with more details about the IoT, and the reasons why I should discuss about it, I feel necessary, at least, to create a connection the IoT technology with a collection of old technology.  My plan is to start a series of mini posts with about 1000-word long on the following subtopics:

  1. What is IoT and Industry 4.0 (this article)
  2. Pros and cons of IoT
  3. Application: from smart applicants and industry
  4. DIY: what is this for tinkers like myself
  5. Smart city: this is my research topic
  6. Security

What is IoT?

According the Wikipedia, Internet of Things (IoT) is:

the  network of [1] physical devices, vehicles, home appliances, and other items [2] embedded with electronics, software, sensors, actuators, and [3] connectivity which enables these things to connect and exchange data, creating opportunities for more direct [4] integration of the physical world into computer-based systems, resulting in efficiency improvements, economic benefits, and reduced human exertions.
  • with the number in bold was my insert to mark the key components:

So yes, IoT is a network where physical objects interfacing with embedded electronics that collects data, filters them by software and algorithm, then controls other clients or those physical objects themselves.  The integration takes advantages of several blocks of technology that otherwise not as effective as those blocks acting alone such as sensor, robots, statistics, Artificial Intelligence, cognitive computation.

I view IoT as a framework or a tool such as Internet where its use and intended goals are directly affected by the creator's intention.  Think IoT as a new technology. It helps improve the productivity when Things are connected and exchanged information.  IoT alone neither beneficial nor harmful, but it is an amplifier, make the outcomes more beneficial or with bad intention more harmful.

A draw of Internet of Things

What is Industry 4.0?

Again, according to Wikipedia,

Industry 4.0 is a name given to the current trend of automation and data exchange in manufacturing technologies. It includes cyber-physical systems, the Internet of things, cloud computing and cognitive computing. Industry 4.0 is commonly referred to as the fourth industrial revolution.

Apparently, Industry 4.0 feels more like an application with the intention to improve productivity either by a better, timely decision or more efficient to use resources.  It puts strong focus with interface of physical objects with electronics and the processing of collected information.  The computing power is at the central of the factory to command/recommend options.  The two interfaces: cyber-physical and purely cognitive/AI computing are connected though IoT so that the data transfer and making decision are almost instant or faster than the changes in physical activity.

Differentiating among levels of advancement the industry is critical for case study and improvement.  Industry 4.0 should be treated as the integration of the automation from the Industry 3.0 and a recent development in AI computing.  

Industry 4.0
Source: wikipedia

I stress this point because I worked in a garment factory for 7 months and the Industry 4.0 was mentioned a bit too much.  That is in part due to the dependence of this industry to the labor skill.  The instability of  work performance leads to the tendency of buffering, or to stock more than one piece of material just in case there is a bottle-neck stage or choking in a station of the production line.  If a process have 30 stations, and each station keeps 3-pieces as buffering, there is (3+1)x30 pieces at a given time, a one 1x30 pieces are in process or is being created value.  With IoT, there is hope that the inventory can be tracked better, or potentially cheaper.  The RFID (radio frequency) tags for its piece of materials are not cheap with a full-scale implementation.  The wireless-ness with a rich data band of IoT is beneficial to integrate a larger scale process with data with enough resolution to analyze and produce meaningful recommendation.

A primer to manufacturing jargon

Without diving too deep into the jargon of manufacturing, there are a three concept related to time: takt time, cycle time, and lead time.

  1. Lead time: the shortest time from a customer orders product to the time that product shipped to the port.  There the lead time for shipping, for pro-processing, production, custom as well.
  2. Cycle time: the time for one process from start to finish. So for one product, the cycle time is the summation the cycle time of each station.
  3. Takt time (pace, rhythm): the pace of time we need to deliver a product based on the customer demand and the time we have.

Why so many types of time, you may ask?  Well, the head of planning department certainly prefers the lead time, the production manager likes to know the takt time and the engineers should know the cycle time. Each data serves different goal posts, and either for projection (lead time), monitoring (takt time) or process improvement (cycle time).


So, really why IoT and manufacturing process now?

For example, a cycle time of a finished piece of garment could take less than a minutes to several minutes depending on the configuration of the production line, which process is bottle-neck and needed more to have a parallel station.  To have the shortest cycle time is not the hardest problem, but to have to both cycle time and takt time as small as possible and the less number of the machines included as well.  Sound like an optimization problem, and I agree. With RFID, counting is easy, but that is the resolution for now. With IoT, the hope is  a sensor integrated to each or key machines for counting with timestamp, possible for visual inspection, and if 'Andon cord' is included, the problem is faster to detect, and therefore, faster to solve.

Wireless communication with sensors and computers are cheaper to a new factory, and retrofitting a old factory as well.  This should be another post for the pros and cons.

Some final thoughts of part 1

So, what are the take home messages for Industry 4.0 here?  First, the Industry 4.0 focuses on integration by connectivity.  It is not only robotics or  automation but they are a part of a bigger process.  Industry 4.0 takes the automation and interface them with autonomous decision guidelines by the computing power at a factory scale, even at the supply-chain scale.  If you are a food processing company, with Industry 4.0, you want to have data from the weather forecasting data with an automatic irrigation, to the sale of the products and customer satisfaction. By the way, those data are no longer acquired by a marking team or a field team, but by satellite images, in-situ sensors, Amazon-liked dash buttons to track the sale, and with social analytics to evaluate the customer feedback.

At the end, the IoT is a method, a tool and served as the conduit between a physical world to a computer power for making decision.  One of difference of IoT conduit is the speed, and multi-direction.