Australian smart city technology firm plans US expansion
March 11, 2015
Australian technology that is empowering smart cities in Australia, Singapore, UAE, India and Europe will be unveiled to US audiences at next week’s GPU Technology Conference in Silicon Valley.
Australian data analytics company SenSen Networks developed the IoT analytics platform based on GPGPU architecture powered by NVidia. The platform is compatible with all major cloud and big data interfaces.
It works by analysing data from sensors such as video cameras (IP, digital, panoramic and analogue) as well as from laser, radar, GPS units and even Kinect devices to produce meaningful outputs for decision makers and owners of assets and infrastructure.
“Our core skill is in using GPU (graphics processor unit) technology to analyse streams of sensor data in real time, recognise all interesting events and create accurate and structured metadata that lead to high-value business insights and operational applications,” said company founder and CEO Subhash Challa. "We saw the potential of GPUs early in their evolution, starting with terabytes of data and now capable of analysing petabytes of data streaming in real time – just what you need to analyse what’s happening in a smart city.”
SenSen’s base in Melbourne, Australia, has helped the company stay under the radar and fine-tune its R&D efforts away from mainstream industry scrutiny. Such humble beginnings suit the 32-strong team, many of whom moved to Australia from countries throughout Asia-Pacific to work with Challa, who is globally recognised and a published author in the fields of data fusion and object tracking.
Five years ago Challa left a comfortable job in academia to start SenSen as a video analytics company building applications for a small but select group of clients. They sought out SenSen because of a common need for a visual record to be used for evidence as well as compliance reasons.
“Our background is in image processing and sensor networks and this means we were able to build a platform that is very efficient and highly parallel for processing large blocks of data,” Challa said. “Our IP is based upon hundreds of efficient algorithms and cores for handling multiple tasks simultaneously – more than 150 man years sit at its heart.”
The company’s blue-chip client list allowed it to roll out and stress-test its products in real-life situations such as law enforcement, traffic infringement, consumer tracking and retail measurement.
About 12 months ago, Challa started thinking seriously about whether an IoT platform to create a truly smart city was possible.
“I noticed that while there were competitors in vertical niches, no one was able to adapt their technology to solve problems other than the one they were focused on,” Challa said. “But everything we learn from one field goes back into the platform to make the entire engine run better. It’s remarkable what you can transfer from one field to another – watching a blackjack player place bets and a shopper looking for a new dress is all a study of human behaviour and decision making, and we analyse how those decisions are made in real time.”
Based on this success, the company has fielded calls from public transport operators, shopping mall managers and fast food franchises eager to know how their customers are likely to behave in the face of competing marketing messages and changes to store layouts. Seeing a wide-open market opportunity, Challa was faced with two major challenges for his company – the platform had to be scalable to cope with the dizzying array of new data sources, and the company itself had to move away from a business model that relied on making direct sales to end users.
The first problem was solved by head-hunting a vice president of engineering from a peer-to-peer games company, and the latter by convincing a COO experienced in building partnerships to leave his own start-up company to join SenSen. Both joined full time at the start of the year.
“We see a global opportunity with our IoT platform thanks to the unique way we are able to gather, store and analyse data,” said Challa. “None of this would have been possible without the advent of GPGPU technology. We have been overwhelmed with interest from sector specialists asking us to build applications that use our analytics engine and platform. As we are unable to pursue all these opportunities ourselves, we have made the engine cloud ready and opened it to other developers to uncover hidden value, reveal fresh insights and discover new patterns. We look forward to seeing what new ideas they come up and working with them to bring those ideas to market.”