BIG DATA BASICS
These are the 5 V's of Big Data
What are Volume, Velocity, Variety, Veracity, and Value?
This maintenance approach uses sensor data and ML to predict failures before they occur.
What is predictive maintenance?
Samsung's Nexplant Analytics platform targets this industry.
What is manufacturing?
Apache Kafka is used for this type of real-time data processing.
What is stream processing?
This document establishes team roles and responsibilities for Assessment 2.
What is the group contract?
Unlike predictive maintenance, this traditional approach waits for equipment to fail.
What is reactive (corrective) maintenance?
These devices collect real-time equipment data to enable predictive maintenance.
What are IoT sensors?
The Sahal et al. (2020) article focuses on these two specific industries for predictive maintenance.
What are railway transportation and wind energy?
This framework processes petabyte-scale data across distributed computing clusters.
What is Apache Hadoop or Apache Spark?
This Assessment 2 criterion is worth the most marks — a whopping 25 points.
What is the project proposal (or solution design)?
This type of analytics tells you what will happen, not just what has happened.
What is predictive analytics?
This Industry 4.0 concept creates a virtual replica of a physical asset to simulate behaviour and predict failures before they happen in the real world.
What is a digital twin?
Siemens monitors over 30,000 of these renewable energy assets globally using big data stream processing to predict blade fatigue and gearbox wear up to 6 weeks in advance.
What are wind turbines?
This cloud architecture places computing power close to the data source for low latency.
What is edge computing?
Assessment 2 requires evaluating this from your Assessment 1 work — the current state of the smart industry app.
What is the current implementation?
Descriptive, diagnostic, and predictive analytics all lead to this most advanced form.
What is prescriptive analytics?
This practice involves automatically updating a predictive maintenance ML model as new failure data arrives, so accuracy improves over time without manual retraining.
What is continuous training (or continuous model retraining)?
This Asian city-state's Mass Rapid Transit system uses big data analytics across 130 km of track and 120+ trains to predict rail degradation and signal failures before service is disrupted.
What is Singapore (MRT)?
This distributed file system, originally inspired by Google's internal storage system, stores petabytes of data across clusters of commodity hardware in Hadoop environments.
What is HDFS (Hadoop Distributed File System)?
This section of Assessment 2 is worth 25 points and requires professional diagrams and schematics.
What is solution design?