Industry 4.0

Industry 4.0 is a buzword much loved by marketing and could easily be dismissed as hype. However, the concepts underpinning it are indeed exciting. More than any time in history, large volumes of data are available to us due to the mass connectivity and advanced sensors of the Internet of Things (IoT), and the technology available to analyse that data and implement intelligent autonomous systems, promise to lead future industry to improved efficiency, profitability and quality of life. The ability to use this data and these technologies are what defines Industry 4.0, the 4th industrial revolution.
Project Topics
The following general topics and project ideas may be used as a guide for proposing an engineering honours project with the University of Tasmania.

Advanced Agriculture

Agricultural environments benefit from significantly distributed Multi-Agent Systems utilising artificial intelligence combined with advanced environmental interaction. Any addition of automation and reduction in dependence on manual labour can improve access to fresh produce and profitability. Project topics include:

  • Irrigation optimisation based on weather prediction, soil moisture, local weather readings and stage of plant growth.
  • Drone automation for monitoring and maintaining crops or livestock.
  • Advanced sensors or image processing for tracking plant growth.
  • Any developments towards fully automated crop production (if you’ve ever played Empire Earth you know the importance of robot farming!).

Smart Homes

Smart Homes incorporate sensors, communication and computational technology for monitoring, automation and control. Goals of the smart home may include residential comfort and safety, healthcare, security and energy conservation. Project topics include:

  • Residential behaviour prediction through machine learning for automation and security.
  • Home incident monitoring through realtime audio analysis.

Cloud Computing

Cloud computing can be applied to any of the other topics discussed here since it involved the advanced processing of data regardless of where that data originates. Projects may be chosen from any cloud computing topic involving the gathering, transmission, processing or distribution of IoT data. For example:

  • Distributed data transfer protocols based on Decentralised Hash Tables (DHT), applicable to agricultural, smart home, personal tracking or smart grid distributed data.
  • Development of a common, distributed protocol suite for the unification of IoT data transmission across industries.

Personal Tracking

Through internet connected personal mobile devices such as mobile phones and smart watches we already share significant volumes of data with services such as traffic monitoring and route optimisation, exercise tracking, or health monitoring. Such services process sensor data including location information, heart rate and activity. Collection of additional types of information, new data processing techniques or new applications and uses of collected data are all appropriate for project topics. For example:

  • Person identification and tracking based on personal behaviour pattern recognition.

Smart Grids

Intelligent power networks provide opportunities for improved power supply through solutions that draw from increased communication and information processing capabilities present in the smart grid. Additionally, the increased monitoring and control within components of the smart grid may be employed to improve observability, controllability and optimality of distributed components such as distributed generators (DG), storage, plugin electric vehicles (PEV), sensors and smart home devices.

Current Projects

iSIM: Intelligent Smart Home Simulator

Davis Allie
Ben Millar
  • Develop a smart home simulation tool (iSIM) suitable for generating datasets of device usage in generic environments and for varying person behaviours.
  • Data realism Synthesised data should have a maximum error of 30%.
  • Data quantity At least 7 days of data with a minute-scale for any environment.
  • Performance The data generation process should be efficient and performant.
  • Ease of use The system should only require the inputs of a home environment and desired inhabitant properties.

Automatic Aeroponic System for Indoor Farming

Owen Tilley
Ben Millar
  • Design and produce an effective indoor aeroponic grow system.
  • Design and produce an interactive control system so users can remotely access the growing space.
  • Develop both of these systems with a cost-efficient and practical mindset.

Residential Behaviour Prediction with a Neural Network

Linxuan Zeng
Ben Millar
  • Monitor residential behaviour.
  • Train a neural network with resident's tracking data.
  • Predict the resident's destination with at least 70% success.
  • Develop a smart light (and potentially other appliances) activated by the neural network.

Residential Behaviour Analysis and Anomaly Detection

Haixu Wang
Ben Millar
  • Monitor residential behaviour through sensors in the residence.
  • Apply an LSTM network to detect abnormal behaviour such as door-opening against the resident's usual pattern.
  • React to abnormal behaviour and initiate safety procedures.

Automatic Garden Watering System with Weather Forecasting

Yibo Wang
Ben Millar
  • Build an automatic irrigation system.
  • Develop a local weather prediction system based on IoT devices.
  • Optimise the irrigation process to maximise water conservation based on the predicted weather.

Maximise Residential Comfort by Controlling Light Levels with Machine Learning

Yifei Jiang
Ben Millar
  • Use light sensors to ascertain the approximate illumination at a location within a room.
  • Determine the optimal light intensity according to the residents' needs.
  • Dynamically adjust the light level to the optimal intensity according to the resident's location.
Apply