master-thesis/1-Introduction/Introduction.tex

\chapter{Introduction}\label{chap:intro}

This chapter lays out an overview of this project and thesis. The reasoning and motivation for exploring the topic will be elaborated, followed by the parameters of the project. Finally, the chapter closes with an overview of the various stages of development of this project, from conception to completion.

\section{Motivation}

\subsection{Transportation of Bulk Material}

It is necessary in several industries, including those of mining and manufacturing, to transport bulk material from one location to another. In mining, it may be sand or gravel. In manufacturing, it may be powdered chemicals\cite{protogerakisInterview2022}.

The transportation of this bulk material typically involves the use of a conveyor belt. These conveyors are specifically designed for the efficient transport of bulk material.

\subsection{Measuring Bulk Material Flow}

This transportation of bulk material flow introduces the need to accurately measure the rate at which the material is flowing. This is essential for various tasks such as keeping track of inventory or for control systems. Knowing when a belt is overloaded, running empty or broken is also an important safety concern. This work deals with the specific challenge of measuring bulk material flow on a conveyor.

\subsection{Conventional Belt Scales}

The conventional method of measuring bulk material flow in use in the industry today is the electronic belt scale---as shown in \autoref{fig:beltscale}. These scales use load cells to translate compression and tension into electrical signals. These signals representing weight may then be converted into measurements of volume.

\begin{figure}[h]
	\centering
    \includegraphics[width=0.75\textwidth]{photographs/beltscale}
    \caption{A conventional electronic belt scale.}
    \label{fig:beltscale}
\end{figure}

These electronic belt scales are robust and proven in the field. However, there are also downsides with this approach.

\begin{enumerate}
\item High unit costs as well as higher retrofitting costs
\item Humidity and moisture content of the material may introduce significant errors which may not be easily compensated
\item Vibration from transport and loading introduces noise into the measurements\cite{tomobe2006}
\end{enumerate}

\section{Aims of this Work}\label{sec:aims}

%\subsection{Research Question}
The central \textbf{research question} that is investigated in this work is:

\textit{How can a cheaper and easier-to-install measurement system for bulk material flow on a conveyor belt be designed?}

\subsection{Use of Commercially-Available Products}\label{sec:useOfCommerciallyAvailableProducts}

As given by the research question above, one of the central parameters is the question of cost. Since the cost of industrial equipment can be far greater than the cost of commercially available products, studying alternatives becomes attractive.
 
As an example, the cost price of the Intel RealSense L515 used in this project was \euro{380}\footnote{Due to Intel announcing that they are discontinuing their LIDAR sensor series, the price of this particular product has risen up to \euro{570} as of January 2022.}, whereas the SICK LM400 used by Fojtik\cite{fojtik2014} can cost upwards of \euro{4000}\footnote{This price is an aggregate estimate based on multiple online merchants as of January 2022}.

\subsection{Use of the LIDAR Sensor}

The usage of the LIDAR sensor was implemented in order to fulfill the second requirement of the research question, namely that the solution must be easier to install than other conventional solutions.

As will be discussed in the following section on design, the LIDAR sensor was selected primarily because it is a contactless sensor. This means that installation can be carried out with little to no adjustments to the existing conveyor belt system. The LIDAR sensor must simply be suitably positioned in order to gather and deliver data.

\subsection{Requirements \& Restrictions}

Besides fulfilling the research question, the design solution should meet the following criteria as well.

\begin{itemize}
\item \textbf{Industrial Robustness} - The final product should be able to withstand the harsh environments that it would likely be installed in, i.e. in a gravel quarry. This means the product must be adequately housed and protected from the environment, against vibrations and shocks.
\item \textbf{Industrial Connectivity} - The product should be able to interface with existing industrial networks, i.e. using Industrial Ethernet.
\item \textbf{Real-Time Ability} - The product should ideally deliver values in Real-Time through the required interface. This means not only a high enough data resolution but also high determinism.
\item \textbf{Remote Control} - The product should be able to be configured and diagnosed remotely, in order to prioritize simplicity of installation and maintenance.
\end{itemize}

\subsection{Conception of the Design}

Analysis of the research question as well as the other requirements have led to the use of the following components to build the final product.
\begin{itemize}
\item \textbf{Raspberry Pi 4 Model B} - Provides a low-cost platform with a Linux kernel and OS for processing data
\item \textbf{netHAT} - A HAT format extension module for the Raspberry Pi that provides Industrial Ethernet capabilities
\item \textbf{Intel RealSense L515 LIDAR Sensor} - Commercially available LIDAR sensor unit, compatible with the Pi
\end{itemize}

More details on each of these components are provided in \autoref{sec:componenets}.

\section{Approach}
The following is a layout of the steps taken in order to realize the final product of this project. For more details on the specifics of the steps, see \autoref{chap:design}.

\subsection{Interfacing with the LIDAR sensor}
The initial step of this project was naturally to establish an interface with the sensor itself. This includes:
\begin{itemize}
\item Preparing the development environment
\item Building the necessary libraries and drivers
\item Developing a basic test software to manipulate sensor data
\end{itemize}

\subsection{Proof-of-Concept Software}
After being able to successfully interface with the sensor, a proof-of-concept software was designed and developed. Later, a GUI interface was also added to the software to improve ease-of-use. The software was designed to be able to do the following things:
\begin{itemize}
\item Display the raw sensor data in a meaningful way
\item Display, calibrate and use the sensor data as a line scanner, with the \textbf{cross-sectional area} as an output
\end{itemize}

\subsection{Laboratory Prototype}
Once the proof-of-concept software was stable, the setup was moved into a laboratory environment in order to further develop the main functionalities of the prototype. Among the functionalities that were developed were:
\begin{itemize}
\item Remote acquisition the raw sensor data over the network
\item Image preparation (offset, rotation, skew)
\item Cross-correlation methods to determine belt velocity
\item Profinet interface to deliver processed data
\end{itemize}

\subsection{Field Testing and Refinement}
Eventually, a stage was reached where development on the prototype in a small-scale laboratory setting was no longer adequate. Development and testing were then continued on-site at a gravel quarry in order to validate laboratory results and further refine the software.