We're almost done with Internet Appreciation Week here at p80. Before we bring it to a close, here's one more serving of food for thought.
Let's start out with a somewhat new, somewhat common, and somewhat confusing idea: the Internet of Things (IoT).
In very basic terms, the Internet of Things means all of your devices will be connected to and sharing data through the Internet. This includes things that were never originally online, like your coffeemaker, your car, and your alarm clock.
You set your alarm for 5:00am. Your alarm clock goes off, and tells your coffeemaker to start brewing and your shower to heat up. Your coffeemaker senses you're getting low on filters, so it orders more on Amazon. Once the shower turns off, your TV turns on and flips to your favorite morning news channel. You hear the coffeemaker beep as you towel off from your shower.
And so on.
Pretty convenient, right? Also, kind of scary. Sounds like a tiny step away from machines coming to life and taking over the world.
Realistically, if we were to connect all of our home devices to the Internet, you would also be able to hook up all of your manufacturing machinery.
Your machines could notify you the second something goes wrong, and immediately order new parts for themselves or be inspected from a distance (among other things).
According to the Harvard Business Review, the IoT provides four different functions for smart, connected machinery:
- Monitoring - constantly takes stock of the machine's condition, environment, and operations. Enables alerts and changes.
- Control - allows for personalized user experience and control of machine functions from afar.
- Optimization - allows for enhanced performance and predictive diagnostics (plus service and repair)
- Autonomy - machine can coordinate itself with its buddies, self-diagnose and service, operate itself, and perform autonomous enhancement and personalization.
Obviously, we haven't fully implemented the Internet of Things. Experts aren't entirely sure when we will reach that point - the current debate is focused on how this could change national and global societies, and how it would affect our day-to-day lives. But, we should expect to see a great deal of IoT progress by 2020.
We should probably expect to see some experimental manufacturing IoT integration in the next few years.
So, what else is in the cards for manufacturing? Here are some up and coming IoT trends within the industry.
1. Industrial Big Data
(Relates to the "monitoring" aspect of IoT)
Regular "Big Data" refers to collecting large amounts of data on human behavior and using it to identify trends and patterns. These data sets and identified trends are used to enhance advertising, predict social movements, block terrorism, and more.
Many people are uncomfortable with the notion of Big Data, as it can be seen as just another way for companies and the government to violate our right to privacy.
On the other hand, Industrial Big Data takes the basic premise of Big Data (collecting large amounts of data to identify patterns) and applies it to manufacturing.
Also known as "data-driven manufacturing," industrial big data relies on the Cloud and other wireless data sharing services to store and utilize manufacturing data.
- Operational Data - according to Forbes, "a company that produces a personal care product can generate 5,000 data samples every 33 milliseconds, resulting in 152,000 samples per second, or 13 billion samples per day, 4 trillion samples per year."
The more data you can gather on your machines and processes, the more confident you are in the quality of those processes and your final product.
- Data Analysis - your machinery will make use of software that allows for predictive analysis. It will take the data it's constantly producing and locate patterns, trends, and possible fail points.
You can then act on these analyses (moving into smart manufacturing, which we'll go over in a second) to optimize your processes and production.
- Data Mobility - lets you check the status of your machines and operations on the go. Your machinery would transmit data to the Cloud or another service, allowing you to access it directly on your phone.
Industrial Big Data brings manufacturers a level of insight that they've never had before. They can then take this data and apply smart manufacturing to improve ALL aspects of the business.
2. Smart Manufacturing
(Relates to IoT's "control" and "optimization")
Smart manufacturing (as opposed to... dumb manufacturing, I suppose) refers to a manufacturing environment where all data, operations, and information along ALL processes are made transparent and actionable via sophisticated technologies.
According to Industry Week, smart manufacturing "gives enterprises full visibility which in turn supports streamlining business processes and optimizing supply and demand."
Smart manufacturing takes advantage of Industrial Big Data and puts the theory into action, improving processes, production, and the bottom line.
Some of the major benefits of smart manufacturing include:
- Using predictive analyses to head off problems before they happen
- Staying ahead of the curve and implementing progressive tech and business tactics
- Providing better service and higher-quality product to customers as quickly as possible
- Complete tech visibility and integration
Some big-name manufacturers have already adopted smart manufacturing: General Electric, Harley-Davidson, and Cisco are just a few.
3. Machine Learning
(Relates to IoT's "automation" function)
According to SAS, machine learning allows computers to find hidden insights without being explicitly programmed where to look.
We're already seeing machine learning creeping into the landscape, in the form of online recommendations and Google's self-driving car.
Machine learning could be a game changer for manufacturers. We've discussed smart manufacturing - optimizing your manufacturing business with insights, guided by a human hand - but smart machinery would almost take human interference out of the equation.
There would be less room for error, faster adjustment to changes made based on predictive analysis, and self-servicing capabilities in the event of a technical problem.
However, as TCS states, "manufacturing operations — such as repairing an aircraft engine, planning the product mix in cement production, or ensuring energy control in a large facility — are still largely dependent on experience-based human decisions."
This is because of a few key implementation challenges:
- Most data scientists have no experience in the manufacturing industry, and so cannot apply machine learning solutions to business problems.
- There is currently no way for manufacturing businesses to oversee and properly implement machine learning processes in day-to-day operations.
- Data captured by the machines may not be relevant to manufacturing processes.
- Currently, there are very few individuals who have the technical knowledge to apply Industrial Big Data to the machine learning algorithms.
As things stand currently, there are quite a few hurdles to clear before we can expect full implementation of machine learning in manufacturing.
Until then, we'll have to rely on the good ol' human touch.
Industrial Big Data, smart manufacturing, machine learning: these are some of the exciting things you can look forward to in manufacturing.
The Internet is slowly intertwining itself with literally everything - manufacturing is no exception. Best be prepared, because it's a "when," not an "if" at this point.
All the more reason to become familiar with how the Internet works.