A Case Study on Using Functional Programming for Internet of Things Applications

In this paper a case study, which shows that the code size and complexity of a system which collects and interprets sensor data in an Internet of Things scenario can be reduced using functional programming techniques, is presented. On one hand this is especially important for security reasons: Such a system must run for a long time without an effective way to distribute software patches. On the other hand in this kind of system the consequences of a malfunction (intended or not) are much more critical than in standard computing situations, because real world buildings or industrial sites are affected. From a high level perspective the data processing at the base station of such a sensor network can be considered as a set of mathematical functions operating on a stream of values. Each function creates a new stream of values, which might be processed by another function. This means that the complete functionality can easily be described and programmed in a functional language, such as elixir, Erlang or Scala.