The invention of IoT devices has exponentially increased data volume in manufacturing systems. Where an earlier network might be handling gigabytes of data, now the devices generate terabytes of data. This volume of data presents challenges in both managing and interpreting the data. Traditional factory cloud options are simply running out of bandwidth to handle the volume. Therefore the trend is toward a hub-and-spoke system of network topography.
Now a lot of data is interpreted and acted upon in a configuration known as edge computing. In this situation, typically Software on a Chip (SOC) architecture ensures that decision-making is in real time and at the node itself rather than centrally controlled through the server. The data and decisions are then transferred asynchronously on the factory cloud. The cloud servers then perform big data–based trend analysis and pattern discovery to allow managers to make more informed decisions.