Altavec
Altavec uses LiDAR and AWS Machine Learning for bushfire prevention
Victorian-based energy network operator AusNet chooses to partner with Altavec to mitigate bushfire risk through proactive vegetation management
By jnally on 31 August, 2022. Original article published on Spatial Source, https://www.spatialsource.com.au/altavec-uses-aws-ml-for-bushfire-prevention/
ltavec has been the trusted partner of Australian utilities for over 20 years. Delivering excellence in geospatial services via advanced geoanalytics, our technology platform empowers best-practice asset management using complex data to provide informed and actionable insights.
Previously known as Geomatic, the new brand name Altavec reflects the provenance of past AusNet ownership and embraces the new ownership by Tech Mahindra.
Victorian-based energy network operator AusNet chooses to partner with Altavec to mitigate bushfire risk through proactive vegetation management. Helping to manage 54,000 kilometres of AusNet powerlines across the state, Altavec uses the machine learning capabilities of Amazon Web Services’ (AWS) SageMaker platform to improve and scale its vegetation classification and analysis process.
The machine learning approach proactively mitigates the risk of bushfires caused by powerlines, ensuring the safe and regular supply of power to business and communities.
Since 2009, Altavec has been capturing LiDAR data. This feature rich data is used to create a 3D model of the entire AusNet network, helping to identify vegetation growth that needs to be trimmed for bushfire safety.
Increasing the automation level of the LiDAR classification process has been an area of development undertaken by Altavec over the past 12 months, with the aim of delivering scale, speed and accuracy improvements.
Partnering with the Amazon Machine Learning Solution Labs team, Altavec and AWS used Amazon SageMaker to build deep-learning models that automate the classification of the LiDAR data in five specific vegetation management categories.
This resulted in 80.53% classification accuracy across these categories, saving AusNet an estimated $500,000 in annual labour costs.
The highly sophisticated and innovative solution significantly reduces the hours spent manually correcting data points by up to 80%, enabling vegetation classification and analysis outcomes to be achieved faster and at a much greater scale.
Nathanael Weldon, Machine Learning Specialist, presented Altavec’s machine learning advancements at the AWS Summit Online Australia and New Zealand in May 2022. Altavec and AWS demonstrated how our advanced technology, built on top of AWS’ technology platform, is making asset and vegetation inspections more effective, safer and cheaper for electricity utilities, and ultimately bringing benefits to the whole community.
Already supporting power companies Ausgrid, Horizon Power and others across Australia, Altavec is preparing to expand its solutions across APAC and the world. Working with energy providers in New Zealand, Europe and the USA, Altavec will provide innovative solutions that keep networks compliant and communities safe in an ever-changing and volatile environment.
Altavec provides end-to-end services for asset assessment including capture of LiDAR and imagery, identification of defects, field management systems including auditing and compliance, digital twins and locational data services, all with the aim of keeping critical infrastructure, customers and communities safe.
We are delivering the digital future to utilities, now.
Contact Altavec at solutions@altavec.com, visit our website altavec.com or call our customer team on 0427 166 551.