Smart Factory: Big Data and Cloud Solutions

In any smart factory systems, huge amount of real-time data are generated.  In order to efficiently and effectively analyze these data, cloud computing must be used over traditional local computing methods.  MtM+’s smart factory solution is cloud ready and this article aim to discuss this in detail.  The following diagram illustrates one particular implementation on a motion detection system.

MtM’s DCS Data Collection Station collects real-time motion (vibration) data from all the machines in operation.  This data is sent to a local server through BLE MESH network.  Traditionally, the data will be stored and analyzed locally.  In the cloud ready solution, the local server doesn’t analyze this data but instead transmit all the real time data to a cloud server over the internet.  The storage and analyze are done via a vast cloud server network.

One might ask, why go through the internet when his can be handled locally.  The following tables look at the pros and cons of both solutions.

For a smart factory, the amount of data generated is enormous.  It requires tremendous processing power to analyze this locally which requires significant investment in initial setup and continual maintenance.  Many companies see the benefits of shared computing power and have setup cloud systems to perform preciously that.  They charge a monthly or yearly fee for their cloud service which includes everything from data download to data analysis.  This is definitely a more cost effective solution for most companies.

The single most concerning issue with a cloud computing solution is security.  Any system that is online is prone to attacks from hackers which can be catastrophic.  A local system is definitely the most secure as everything is kept locally but one must weigh between cost and security.   MtM+’s current cloud solution is setup with the Alibaba Cloud (Aliyuan).  This is highly secure and is outlined in details in the following link:    The following diagram illustrates preciously MtM+’s implementation.

The vibration sensors detect real-time data where the Data Collection Stations (DCS) transmit this information to local server/gateway through either Bluetooth Low Power (BLE) or LoRa interfaces.  Once the gateway successfully collect these data on the local server, the on-premise Dashboard will display these in detail before uploading them onto the cloud server (in this case, Aliyuan is used).  Once uploaded, the information can be accessed and analyzed anywhere with an internet connection.   The following figure shows two different dashboards, the left being the local server and the right being the cloud server.

Looking at our sample case of vibration detection, the local (on-premise) server can generate status reports and display real time data. This data is also coarsely analyzed to detect abnormal activities (ie power outages). While these features are available for both local and cloud servers, local server do offer some additional services. An on-premise SQL database on the local server acts as a backup to the generated data. If needed, the DCS devices can be reprogrammed or modified locally but not through the cloud.

The cloud servers however offer very powerful computing power which would be very expensive to implement for any company. With the cloud, these big data can be analyzed effectively and efficiently with any combination of parameter matrices as shown in the following diagram.

Provided the advanced security of most available cloud service providers today, the benefits of their advance analytical tools make cloud computing a vital part of any smart factory implementation. For any question and inquiry, please email MtM+ at