The Value of IoT
by Andrea Dorn
Everywhere, for the last few years, you are faced with the Buzz Word "IoT" - Internet of Things. I think it was in 2014 at the Embedded World: I got the first time in contact with it. And I asked what is "IoT" - what the hype is about? Somebody told to me: "Well, that means that everything is now connected. This causes new challenges and we have the solution for those challenges." I watched the situation, followed lectures and I realized, that the real value of IoT is not the connection - that's a matter of course.
The real value is the data.
Or lets use another Buzz Word: "Big Data". Amounts of data which are big, complex and which change fast.
In Germany there is an additional relevant Buzz Word existing: "industrie 4.0". This is a politically stamped word. Our Chancellor Angela Merkel puts emphasize on the digitalization of industrial production. In the following context IoT and industrie 4.0 can be seen equally.
In my spare time I like riding my small, wild Iceland pony. While I was thinking about the value of "The new World - IoT", I wanted to ride across a farm. Here in southern Germany, most of the Farms have cows. Milk Cows. I wanted to pass the door to the cowshed. But my pony didn't want to. There were some strange noises. It sounded like compressed air or similar. I watched through the door and saw a fully automated milking and feeding machine. I was fascinated - quite to the contrary to my horse, which refused to come with me to that door. So I stood there, laterally, one arm far back so that I could watch the cows while holding tight my pony. It must have been a funny picture. The system I saw was called "Herd Manager". It takes care for an automated routing so that all the animals of a herd are milked, get the right food or are sent to the willow. Moreover it identifies cows which need to be separated for some reason. The farmer can see on a screen - in real time - which cow has been separated and why. They call it a clever system. But what makes the system so clever?
The Data - respectively the ability to use and handle this data. So how does that work? The amounts of data has to be stored somewhere, and it has to be analysed somehow - in real time. So that the clever system - here the "Herd Manager" - can act as expected. And that's a real challenge, isn't it?
How to deal with Data is not a new challenge.
In the 1960s a concept was created to manage data in a separated software layer between the operating system and the application. So that copying, mixing and restructing of files was automated and did not belong to the daily work anymore. One of the first big Data Management Systems was IMS (an informationsystem from IBM) with the language DL/I (Data Language One) with a hierarchical structure. In parallel CODASYL defined a network-like structured database. Edgar F. Codd developed the foundations of the first experimental database system. The relational database management systems replaced in the 1980s the hierarchical and network-like systems. In the 2000s the open source database management systems gained prominence especially MySQL.
In meantime there is a hughe number of database management systems available (open and commercial). They are used in many different (sometimes mission critical) systems from consumer electronics, medical devices, robotics, industrial automation, solar systems and many other markets.
Database and Relational Database Management Systems
Everyone knows databases, everyone uses them - probably even daily. But how the thing works is not part of the general knowledge. Reason enough to look once behind things and to deal with the basic of the database. A relational database is a collection of objects (data). These data are placed in tables. A data object can be a statement, a link or just a text. A database can have a number of tables. The tables in which the data is stored are represented by columns and rows. A complete line of a database table is considered a record.
The number of columns is here responsible for the length of the record. A field is a specific column of a record. A Relational Database Management System (RDBMS) provides different tools to create one or more databases, to fill them with data and to manage the data. A RDBMS has typically at least one user interface and can manage one or more databases. Each user interface offers the user an opportunity to directly receive commands. These are sent to the RDBMS and executed in the current database. The used language is in the most cases Standard Query Language (SQL). SQL is quite easy to learn - at least the basics. The more complex the search request to the database is - the more complex the SQL-queries are.
Side Information: In other kinds of databases, data is structured as graphs, key-value pairs, documents or other ways. For several years these systems have been gaining popularity (under the Buzz Word NoSQL) in specialized use cases.
The basic requirements are usually two fundamental demands: Avoiding redundancy and ensuring maximum data integrity.
- Avoiding Redundancy: Information should not be stored repeatedly.
- Data Integrity: You can enforce data integrity e.g. in a field for a birthday date you only allow to enter numbers. Moreover data integrity means that in cases like a power failure a RDBMS must ensure that data remains consistent.
A quick access to the data, fast import of large volume of data and the ability to keep records organized for decades... are consequences of the above explained requirements.
Does the world of IoT and Big Data effect those fundamental requirements? The fundamental requirements are still valid but I think the fact that everything is connected confronts us with additional challenges:
- Concurrency - that means that concurrent accesses to a database must be possible efficiently and without blocking or endagering the consistency of the database.
- Data Security - Data are worth protecting goods. Access to these should be limited and controlled. Moreover a lack in the security can also affect the behavior of a system: the safety.
The real value of the future is the profitable exploitation of data.
The handling of data is becoming more and more complex and affects the behavior of a system also. Databases in general can be a helpful tool to represent and handle such situations. I think whether this is an open source solution or a purchased solution remains entirely up to you. It is important to be aware of the mentioned requirements and of the value of the data. Having data under control starts in your own company and may influence the success of your company in the future.
What do you think is the value of IoT? I am looking forward to your comments.
For an automatic notification on new blog articles, just register your EMail address.