The Victorian Government recognises the benefits of innovation to the economy, the community and the environment.

In line with the recommendations of the Victorian Bushfires Royal Commission, the Victorian Government's Powerline Bushfire Safety Program (PBSP) is working to reduce the risk of bushfires started by powerlines.

By investing in world-first research studies into understanding the relationship of bushfires between certain vegetation species and powerlines in bushfire prone areas, vegetation conduction ignitions have been identified through Fault Signature research.

Findings from this research have been made publicly available in order to foster further research, innovative solutions and business enterprise of improved detection equipment and technologies to prevent bushfires from powerlines.

The full authorised data set including 300GB of photos, videos, test logs and report, are available on the DataVic website.

If there is the ability to detect species of vegetation touching a powerline, this presents the opportunity to reduce bushfire risk.

The objective of the 2017 Vegetation Detection Challenge (the Challenge) was to develop an algorithm and concept that can identify what particular plant species cause a fault when touching a 22kV SWER powerline.

The Vegetation Detection Challenge focused on three particular plant species:

  • Salix Species (Willow) - high fire probability
  • Franxinus Angustifolia (Desert Ash) - medium fire probability
  • Schinus Molle (Peppercorn) - low fire probability

A consolidation of the fault signature data for the above three species is also available on the DataVic website - Vegetation Deteection Challenge data.

Algorithm image

The Challenge

There is currently no mechanism that can detect vegetation faults on powerlines in a timely manner.

People in the fields of science, technology and an interest in data were offered the chance to enter a competition to come up with ideas on how to use vegetation fault signature data to improve powerline safety.

Available was comprehensive research data of electrical characteristics of powerline faults that create a fire risk link to vegetation (see above).

The Challenge was launched on 2 May 2017, with individual sprint and check-up sessions made available with final presentations occurring in October 2017.

Competition Final

From a shortlist of four scientific teams, the winner of the Challenge was announced on 1 December 2017 to Melbourne-based mechanical engineering and computer science team, Yidan Shang and Nan Li.

A one month innovation period has been made available to all teams to potentially work with a company towards developing a new product, process or adaptation of technology to identify plant species on powerlines allowing for a response to a fault before it becomes a fire.

Read the summary sheets developed by the teams below and to request the algorithms, email fault.signature@delwp.vic.cov.au

The participant summary sheets and the Challenge video are also available on the DataVic website.

Disclaimer

This material is provided as information only. The Victorian Government and this agency (Department of Environment, Land, Water and Planning):

  1. Makes no warranty, either express or implied, in respect of the material, including (but not limited to) in relation to the accuracy, reliability, availability, or the currency of the material at any time, or as to its suitability for any purpose; and  
  2. Does not accept any liability to any person for the material, or for the information or advice provided on this website or incorporated into it by reference (or for the use of such material, information or advice). No responsibility is taken for any information or services that may appear on any linked websites.

Please note that the relevant team are the owners of intellectual property in the algorithm, which has been licensed to the State in accordance with the terms and conditions of participation in the Vegetation Detection Challenge.

If you wish to reproduce, publish, communicate to the public, adapt, modify or otherwise use the algorithm, you must first contact the relevant team.