Aerodyne taps AWS to power drone platform

Malaysia-based startup Aerodyne is tapping Amazon Web Services (AWS) to power the Dronos software-as-a-service (SaaS) platform that enables drone operators to onboard, analyse and make sense of drone data in a variety of use cases.

Besides building a data lake using the Amazon S3 storage service to store and turn drone data, such as satellite images and weather data, into insights, Aerodyne is also using Amazon SageMaker to build, train and deploy machine learning models to analyse the data.

Kamarul A Muhamed, founder and group CEO at Aerodyne, said the company has been able to help farmers increase their yield by over 60% by using drones for agriculture seeding, plant analysis and mapping, as well as reduce operational costs for customers that operate and maintain critical infrastructure such as telco assets and gas pipelines.

For example, Aerodyne has helped Malaysian energy giant Petronas to digitise its maintenance operations, so instead of flying an aircraft or driving along 2,000km of gas pipelines, smart drones can be deployed to investigate anomalies picked up by sensors.

“These are drones that have AI [artificial intelligence] at the edge that can detect risks and make assessments on whether something is a false alarm or risk that needs to be escalated,” Muhamed said on the sidelines of AWS re:Invent 2023 in Las Vegas this week.

In the telco space, the startup – which has been doubling its revenue annually since its inception in 2014 – claimed to have reduced cellular tower operational costs by an average of 20% and alleviated the need to have workers physically inspect towers. Muhamed said the largest telco company in Australia is already using its platform and that Aerodyne is looking to onboard more customers in the near term.

“Working with AWS has significantly transformed our ability to resolve complex industrial challenges, expand to more countries and deepen our footprint in the global drone community,” said Muhamed. “Through the agility of the cloud and the use of machine learning, we can bring valuable data together to help people across agriculture, telecommunications and energy industries make faster and better decisions about their assets on the ground.”

Aerodyne has a team of over 1,000 drone professionals in 45 countries, so ensuring its platform can scale to meet the requirements of customers in different countries is key. That means instead of developing bespoke services for customers, which it did in its initial years, it has identified common capabilities that customers can deploy.

“We provide a dashboard for customers to choose what matters to them,” said Muhamed. “For example, in thermal inspection, which enables you to predict when certain equipment will fail, some customers may look at differences in temperature while others may want the absolute temperature. We’re enabling all possibilities within our software, so rather than doing things bespoke to anyone, customers can choose how they want to do it.”

It also offers application programming interfaces its customers can use to integrate the Dronos platform with existing enterprise resource planning (ERP) systems, such as those from SAP and Oracle. “Our customers do not want their data to be an island by itself – they want the data to be integrated with their existing ERP and that’s our key focus as well.”

Looking ahead, Aerodyne will be experimenting with AWS’s GenAI capabilities to build a large language model that can help its customers better plan drone flights and visualise close to 1 PB of drone data. This includes insights from digital twins that can help to centralise management of physical assets by combining images and environmental data in real time.

Amid concerns over data security, which Muhamed noted was a “big issue” for customers, working with AWS has enabled Aerodyne to assure its customers its platform is running in a secure environment and that “all the check boxes are ticked in terms of cyber security and safety”.