Innovation Central
Summer Internships

Applications are now closed. Thank you for your interest.

Find out more

Did you know it’s possible to gain real-world experience with a global firm without even leaving campus?
What about being on the cutting edge of innovation in developing impactful business solutions?

ICP’s Summer Intern Program is open to all WA undergraduate university students looking to extend their discipline knowledge with hands-on experience.

This 12 week full time program gives students from all academic backgrounds the opportunity to complete a real-world project put forward by one of our industry clients. Students will develop their project management, planning and design thinking skills via exciting projects involving the Internet of Things (IoT), network technologies and data science.

Interns are supervised by leading researchers from Curtin University and sponsoring industry organisations. Projects range across a broad spectrum of industries, including health, agriculture, local government, transport, infrastructure and more.

Innovation Central Perth acknowledges the importance of diversity and inclusion and we encourage all interested students from all academic backgrounds to apply.

Information for students

Students studying at a WA based University are eligible to participate in the Summer Program of the ICP Innovation Internships.
Internships will run for 12 weeks during the period November 2019 through to February 2020 and you will receive a tax-free scholarship of $6000 to support you across the duration of the project. Applicants must not have completed an ICP Internship in the past.

As a condition of the award, you will be expected to:
- Attend a one-week induction in late November/early December
- Make satisfactory progress across the internship period
- Write a project report
- Create a podcast describing your work
- Give an oral presentation to industry leaders at the end of the internship

Awarding of internships will be assessed on each individual student’s merit. We will also be aiming for diversity across the internship program selection.
Students should nominate their preferred projects in their application. If none of the listed projects are suitable, we may consider alternative development ideas subject to alignment with ICP themes, identification of an appropriate academic supervisor and linkage with one or more of our industry clients.

Projects

The available projects are outlined below. You'll have the chance to nominate your top three projects during the application process.

  • 1. Geo Tracking of Hospital Patients.
    When a patient is admitted to a WA hospital, they are likely to attend different parts of the facility to receive care. In some cases, the primary party responsible for the patient can lose track of their location, which can also pose an issue for visiting family. New developments in GPS tracking and geo tagging offer options to monitor the location of individuals moving through complex facilities. WA Country Health and Frontier SI would like to explore this to assist caregivers in optimising their involvement with patients and to assist family in tracking their loved ones during visitation.

    Project Challenge: Design and demonstrate a geo tagging track-and-trace solution to help people in a central location identify where a patient is at any given time and follow their care plan journey to optimise their time at hospital.

    Indicative academic background: Science & Engineering, Application Development

    Vertical: Health

  • 2. Passive Monitoring of Burn Patients.
    Patients who do not have their sleep interrupted to check vital signs experience better recovery post burn injury. Better recovery post burn is defined by faster time to heal, less rates of infection and better health related quality of life than those in the control group. The Fiona Wood Foundation is looking to undertake a research project, to confirm the hypothesis that non-intrusive monitoring of burns patients will lead to better recovery outcomes.

    Deliverable: Design, develop and demonstrate a sensor array for deployment in burns recovery rooms. Consider sensors worn by patient and in close proximity to patient and transmit to a fixed external monitoring site.

    Indicative academic background: Science & Engineering, Mechatronics, Data Science

    Vertical: Health

  • 3. Image Classification of Burn Wound Images.
    Photogrammetry and remote sensing have the potential to support the assessment of burn wounds. In the initial project a simple k-mean clustering algorithm was applied on burn images which were downloaded from the internet. In the meantime, it was possible to capture burn data using the VectraH1 camera. This camera enables the 3D modelling of objects including rendering the model using RGB information. The aim is to use an image classification algorithm (such as SVM), to train the algorithms with the data from one patient and to then apply the trained classifier to another patient’s data.

    Deliverable: 1. Applying colour correction to the original image data

    2. Re-calculating of the rendered 3D models

    3. Applying image classification, trialling different approaches

    4. Rendering of the 3D models with the resulting label images

    Indicative academic background: Science & Engineering, Image Classification, Image Calibration

    Vertical: Health

  • 4. 3D Imaging Database for Clinicians.
    Cliniface is a 3D facial analysis hub that enables collaboration with clinicians, researchers and computer scientists to advance understanding of facial characteristics and their relationship with rare disease and other conditions. Part of the Cliniface project is the capture of 3D faces of patients, the metrics of which needs to be stored alongside personal information in a suitable database. The existing data collection contains information from multiple studies which utilised 3D imaging, however, the current data is saved across multiple spreadsheets, has become unwieldly and poses several issues, including no linkage between the spreadsheets and the 3D image files.

    Deliverable: Set up a functional database that houses patient information, along with 3D metrics derived from 3D face imagery, which can easily be used for data mining by the project partners.

    Indicative academic background: Business & Law, Humanities, Marketing

    Vertical: Health

  • 5. Commercial Movement Analysis.
    Local Councils are exploring ways in which phone location data can be analysed to understand traffic movements to and from industrial areas. The data findings can be used to improve flow and efficiency, as well as any opportunities for improvement and underutilised roads

    Deliverable: Investigate and design a viable data collection program, using Bluetooth based technology, to track individual vehicle movements across key transport corridors into and out of the Canning Vale area.

    Indicative academic background: Science & Engineering, Data Science, Machine Learning.

    Vertical: Local Government

  • 6. Smart Fleet Management.
    Local councils are looking at ways to collect origin and destination data from transport operators which can be used to provide a live heatmap of projected traffic conditions at any given time of the day. This will include data from participating transport operators and Local Government Authorities on planned roadworks and live hazards.

    Deliverable: Investigate and design a viable data collection program, involving transport operators, to collect specified data and then present this for analysis.

    Indicative academic background: Science & Engineering, Data Science

    Vertical: Local Government

  • 7. Remote Sensing for Ecosystem Health & Function on the Wheatbelt.
    Carbon Farming is an emerging industry in Australia and globally, as there is a need for quick, cost effective mechanisms to monitor ecosystem health at landscape scales. The requirement is to inform management practices and identify stand out management in season and progress over decadal timeframes. Carbon Sync are looking to determine how remote sensing data (via Open Data Cube) can be used to communicate change that has occurred with regenerative agriculture practices.

    Deliverable: Deliver 3 case studies of farms in the wheatbelt of WA who have been practicing regenerative agriculture for 10+ years. Utilise data found within the Open Data Cube to monitor the health of these ecosystems and deliver creative reports and visualisations of the changes.

    Indicative academic background: Science & Engineering, Business & Law, Graphic Design.

    Vertical: Agriculture

  • 8. IoT Connectivity for Roborigger
    Roborigger is a load controlling device that allows crane lifting operations in construction and logistics to be performed without the need for people to be in the vicinity of the load. TENSA Equipment, in conjunction with development partners, Multiplex and Woodside, are seeking support to improve Roborigger’s IoT implementation.

    Deliverable:

    Part A: Use deep learning or machine learning to develop a model that can recognise the type of load being lifted and use QR Code/barcode/OCR/RFID to retrieve additional information about the load.

    Part B:. Develop software that analyses on-board sensor data to automatically identify events of interest in real time.

    Indicative academic background: Science & Engineering, Computer Vision, Machine Learning, Software Development

    Vertical: Construction

  • 9. 3D Automating Subsea Inspection Data
    In the majority of subsea inspection projects, the visual data is voice annotated with current status of the infrastructure and anomalies detected. This voice annotation is completed by highly skilled inspection engineers and at great cost to the business. Unfortunately, this methodology is flawed and presents challenges, including the need for filtering and navigating video files to review historical data. Fugro, the world’s leading Geo-data specialist, are looking at ways to automate their subsea inspection data processes.

    Deliverable:

    Part A: Automatically extract relevant information (infrastructure and anomalies) from voice annotated visual data of subsea inspection projects.

    Part B:. Use the above data to create a training dataset of relevant targets for supervised machine learning, to enable automatic subsea inspection and detection of relevant target details.

    Indicative academic background: Science & Engineering, Computer Science, Machine Learning

    Vertical: Oil & gas

  • 10. Using Social Media Data for Space Utilisation.
    Local Councils are looking at different ways to track and record the use of public spaces that they administer. This can be difficult using traditional CCTV or video capture due to privacy issues. Many people volunteer images of their involvement and activities via social media and this could be analysed, to compile data on the utilisation of specific spaces and locations, especially outdoors.

    Deliverable:

    Part A: Leverage Social Media platforms to compile data on activities for specified public locations.

    Part B: Use machine learning to train image analysis to provide a statistical summary of activities.

    Indicative academic background: Science & Engineering, Data Science, Business & Law

    Vertical: Local Government

  • 11. Canning Sustainability Neighbourhoods.
    The City of Canning will select a ‘neighbourhood’, likely to encompass one street initially as a pilot, and support that neighbourhood to implement a number of sustainability initiatives (largely centred on energy management / efficiency, renewable energy installations, water conservation and waste minimisation / circular economy) and then measure the aggregated reductions / impacts of that neighbourhood over time.

    Deliverable: Support the development of the program by pulling together the technologies, monitoring and outcomes measurement tools to create a dashboard that can demonstrate both individual and aggregated outcomes on a shared leaderboard. Also, support the community engagement around the project, which would include working with the homeowners / participants to onboard them into the program and support them to implement the range of sustainability measures that might achieve impact reductions.

    Indicative academic background: Science & Engineering, Humanities, Business & Law

    Vertical: Local Government

  • 12. WesCEF Safety Data Analytics.
    WesCEF maintains a number of disparate systems/datasets for recording and reporting on incidents & hazards, HR, training and other relevant safety performance indicators. The abstract nature of the information and datasets makes it difficult to undertake a cross-platform analysis to gain actionable safety insights. WesCEF wish to better understand the factors that drive their safety performance to inform decision making with regards to delivering better safety outcomes for our workforce

    Deliverable: • Formalise a series of safety Hypotheses for testing. Examples include, but are not limited to, employee tenure vs. injurie rate, level of training vs. incidents, high risk worker profiles and at-risk worker demographics. • Identify relevant systems and datasets for analysis, cleanse and link data as required to enable exploratory analysis and develop a data model to enable interrogation. • Present WesCEF senior HSE leadership with validated, actionable business insights regarding factors influencing safety performance and present findings within a platform that will allow WesCEF to interpret analysis and further expand model in future.

    Indicative academic background: Science & Engineering, Data Science, Statistical Modelling, Machine Learning, Computer Science

    Vertical: Chemicals

  • 13. IdeaLab.
    There is currently a lot of interest in Western Power around developing innovation to help them ‘in the field’ with adding business value. The IdeaLab frames this opportunity into an initiative. Employees will pitch problems to each other in small teams, then determine if the idea is worth exploring, pull a sprint team together, then get to work on progressing the idea. Here is where they build their knowledge and practice of design thinking and agile mindsets.

    Deliverable: Come up with research, summary and recommendations as to how Western Power can make IdeaLab a success – Success meaning that the corporate culture accepts and embraces the initiative to deliver business value on an ongoing basis. (e.g What tools and mechanisms are needed to boost engagement and influence culture? How do employees better engage with Idea Lab? How does IdeaLab leverage existing tools within Western Power

    Indicative academic background: Business & Law, Humanities, Science & Engineering

    Vertical: Utilities

  • 14. Upside Risk Model.
    Western Power are well practised at risk modelling. Traditionally, a utility project will illustrate its committed delivery benefits as part of its management plan, and as it illustrates its committed risks. The flipside of risk is the opportunity or ‘upside’, however in the utility space there is no mechanism for measuring that upside. If there was such a mechanism, it would help expose opportunities to enable decision makers in the organisation to broaden their scope for potentially enterprise value building investments.

    Deliverable: Build an upside risk model for measuring the opportunities of project level risks. • Build a prototype for an upside risk model • Demonstrate how the model can test assumptions through asking questions • The model would need to have a measure or ability to weight probabilities of differing decisions

    Indicative academic background: Business & Law, Humanities

    Vertical: Utilities

  • 15. Bar Code Challenge.
    Western Power’s Kalgoorlie depot have identified a challenge with regard to effectively maintaining different asset classes across their network. Many assets of similar classes have different maintenance requirements due to their age, new technology/component settings, the brand design criteria, warranties, safety standards and so on. This challenge means it’s often complicated and time consuming to manage these assets. They would like to find a simple way, for example through a QR code reader to readily understand what asset class and criteria relates so the asset they are maintaining on the field, to service it correctly

    Deliverable: • Deliver a prototype of a QR coding system or similar that can help readily track the asset information critical to the maintenance of that asset on the field. • Evaluate the cybersecurity challenges with a QR code and how this type of technology could be on boarded into the Western Power IT network.

    Indicative academic background: Business & Law, Humanities, Science & Engineering

    Vertical: Utilities

  • 16. Technology Driven Spotted Quoll Identification.
    Consolidated Minerals currently have a network of cameras distributed across their site that monitor/track native species ‘Spotted Quolls’ moving in and around the mine site and processing plant. They would like to build data to illustrate the quoll’s habitat, and that their habitat is not impacted by mining, and that coexistence is possible. The way the camera works is that there is bait near the camera, the quoll eats, then disappears. This is very rudimentary. - Tracking is ‘manual’ through sensor driven spy cameras - Enticing via baiting is not ideal as it’s not ‘natural’ - There is no way to individually identify the quolls to understand their unique habitat - Due to the network and connectivity restrictions across the site (and associated costs) it’s hard to build a camera network that’s monitored and recorded remotely

    Deliverable:

    Part A: Use the current image data available from the cameras to train algorithms to use facial recognition technology to pick up the unique pattern identifiers on the Quoll’s backs. 1.Dot pattern recognition – develop machine learning algorithms to identify Quolls from pictures already taken 2.Use existing data to train the deep learning needed 3.Potential to identify new Quolls entering the area

    Part B: Use a LoraWAN network to transport the sensor driven pictures through the network, upload to the server, utilise facial recognition technology to identify the Quolls in real time, and based on those identifiers, track their location and movements

    Indicative academic background: Data Science, Machine Learning, Science & Engineering

    Vertical: Mining

  • 17. Power Reduction Through Microgrid Trial.
    Consolidated Minerals have large energy costs on their mine site due to the heat and humidity in the region. They believe they have the opportunity to reduce costs of diesel gas by 50% if they can control air-conditioning in a way that would manage the aircon to stay at specific stable levels to manage humidity and minimise mould creation. They are currently doing research to establish which providers are out there for edge devices/sensor gateways that are technology agnostic, allowing multiple sensors to drive data through to a connectivity platform. Following the trial of air-conditioning control, the data collection work could grow to collect temperature, vibration, air pressure and more, providing a useful diagnostic and control opportunity for Consolidated Minerals to optimise not only energy use but preventative maintenance optimisation.

    Deliverable: Conduct initial research as to what technology/SMEs are out there that could enable this type of edge connection opportunity firstly utilising air conditioners. Research what tech is out there, what has been done before and what industry leaders are in this space. Engage with one tech provider for a trial to collect data to better understand usage patterns

    Indicative academic background: Business/Commerce, Humanities, Science & Engineering

    Vertical: Mining

  • 18. Visitor number estimation in the City of Fremantle
    The City of Fremantle would like to improve its understanding of the total number of people who utilise the Fremantle city centre on an average week and weekend day. The project will provide an average daily population estimate that takes into account the wide variety of users of the city centre

    Deliverable: Using location intelligence and wireless presence tracking, the project aims to: 1. Establish an estimation models to determine the number of visitors and the type of visitors in the city centre of the City of Fremantle 2. Provide average daily population estimates for the following users of the city centre on weekdays and weekend days. (Users include Residents, Local workers, Students, Daytrip - intrastate, interstate and international visitors, Non-day-trip visitors,)

    Indicative academic background: Science & Engineering, Data Science

    Vertical: Local Government

How to apply

  • 1. Download

    Grab a copy of the application pack here

  • 2. Nominate your projects

    Fill out the form, and don't forget to indicate the top three projects you'd like to work on

  • 3. Submit

    Email us your application, CV and academic transcript by
    27September 2019

Submit now

Connect with us

here

Curtin logo Cisco logo Data61 logo Woodside logo