Recently receiving US$2m in seed funding, Hong Kong-based startup viAct is aiming to use artificial intelligence (AI) to create a “zero risk” smart construction site.
Founded in 2016 and based in Hong Kong, viAct has developed an artificial intelligence (AI)-based cloud platform that provides construction sites with round-the-clock visual monitoring for quality, security, safety and productivity management purposes. The application combines computer vision, edge devices and a mobile app. Computer vision detects potential safety hazards, construction progress, and the location of machinery and materials. Real-time alerts are sent to a mobile app with a simple interface, designed for rapid assimilation of information by users in noisy and dynamic working environments.
In March 2021, the startup announced (TechCrunch) that it had raised US$2 million in seed funding, in a round co-led by SOSV and Vectr Ventures, with participation from Alibaba Hong Kong Entrepreneurs Fund, Artesian Ventures and ParticleX. By this time, viAct had more than 30 construction industry clients in Asia and Europe. Current customers include BTP, Swire Properties and China State Construction.
It plans to use its latest new funding for research and development, product development and expanding into neighbouring southeast Asian countries. These include Indonesia and Vietnam, both seen as growth opportunities over the next 5-10 years, with their governments planning for smart cities and new infrastructure. Singapore is also a target as it has mature developer customers willing to adopt AI-based technologies, as is Japan and Australia. Longer-term, the company aims to expand into Europe, encouraged by early adoption of its technology by firms from Italy and Spain
The viAct story
The company was founded by Gary Ng, CEO – right, and Hugo Cheuk, now chief operating officer. Both graduated in building engineering from Hong Kong Polytechnic University in 2002 (see also this January 2021 StartUp.Info interview). Ng then joined an engineering consultancy firm before making a switch to technology businesses (he had senior roles at two Israeli-backed companies: Stratasys and EFI Optitex). Cheuk worked in range of media and manufacturing firms, including a major Asian cosmetics business.
In 2016, the two friends decided to start their own business, combining their engineering and technology experiences, and after identifying construction enterprises as ripe for technology adoption, began testing AI monitoring solutions for construction in 2019. Company-bespoke AI models were too expensive and time-consuming to develop, so they focused on building a standardised AI cloud platform that construction companies can deploy in minutes. Ng told Extranet Evolution that the viAct application analyses video outputs from existing site cameras (usually static devices, but also from drones, tablets and smartphones), using image recognition algorithms built upon a growing database of tens of millions of construction site images – captured during over 30,000 hours of construction activity.
Interestingly, expansion of the viAct team was not hampered by the COVID-19 pandemic. With employees forced to work from home locations, viAct adopted a what Ng called a “remote first” culture. This also enabled it to recruit the best staff regardless of location (it now has some 40 staff, including people working from India, Thailand, the Philippines, even Argentina). And they also apply a more international perspective to viAct’s work.
In January 2020, during the early days of the COVID-19 pandemic in Asia, viAct was the first Asian company to develop a face-mask detection solution for construction sites. Speaking about the latest funding round, SOSV partner and Chinaccelerator managing director Oscar Ramos said:
“COVID has accelerated digital transformation and traditional industries like construction are going through an even faster process of transformation that is critical for survival. The viAct team has not only created a product that drives value for the industry but has also been able to earn the trust of their customers and accelerate adoption.”
How viAct works
Depending on use cases, viAct can capture video and image data for real-time processing or post processing. Captured data is streamed directly to AI engines for data analysis and to generate alerts – typically sent within a minute. Warnings help workers to protect themselves. viAct also helps safety managers or site agents monitor onsite incidents, while allowing senior management oversight of multiple project sites.
Anyone without AI knowledge and coding background can set up the viAct system within five minutes. Its simple to use, highly flexible, plug & play technology makes it easy to onboard new users, says Ng.
To date, viAct has developed over 50 AI modules relating to a wide range of construction environments, plant and equipment, materials and processes. Some are generic (safety modules relating to wearing of personal protective equipment, PPE, for example), while others are specific to particular environments (viAct has modules developed for deployment in tunnels during infrastructure projects, for instance). viAct aims to eventually have a library of AI modules for most processes and working environments. Ng says customers can see how many workers have been deployed, what equipment is operating, the location of materials and tools, etc – many activities which used to rely on human monitoring can now be automated.
viAct sells just its cloud platform, to which customers connect their existing cameras and online systems. As a cloud service, viAct customers pay according to the volume of images monitored by the system’s modules. Cheuk, below right, says:
“Some projects start with feeds from just two or three cameras and use just two or three modules. A basic deployment at that level for six months will cost under US$5,000 – less than £1,000/month. More cameras and more modules will naturally increase customers’ use of our AI service and increase the cost pro rata, but the cost benefits in improved safety, quality and productivity usually mean a rapid return on investment.”
Cheuk adds that, once deployed, the system can be constantly updated.
“Our software has a drag-and-drop interface that allows users to quickly select the modules they want to use, and they can swap between modules at any time – vital as dynamic projects progress, deploy different equipment and manage different processes.”
The company featured in CEMEX Ventures’ TOP 50 most promising startups of 2020. It is also an Autodesk development partner, and has created an integration between its AI platform and Autodesk BIM 360.
The competitive space
viAct has few local competitors in southeast Asia. Its major international competitor is a Boston, US-based firm, SmartVid.io, founded by Josh Kanner (who co-founded Vela Systems – acquired by Autodesk in 2012, TechCrunch, to form the basis of Autodesk’s BIM 360 field mobile application), and also employing former ConstructWare and CTSpace executive Gary Greenberger (April 2008 post).
Also exploiting construction site video content using AI is Dublin, Ireland-based Evercam (April 2020 post: Evercam : site imagery, AI and BIM), which deploys AI image recognition tools to help construction site managers monitor vehicle movements, and, in a COVID-19 context, people movements.
However, viAct is located in one of the fastest growing global construction markets, and has deep local and domain-specific experience – and, perhaps more important – it also has huge volumes of data. Moreover, its business model is geographically very agile, not reliant on existing technologies, and its AI modules seem highly customisable to different corporate and country-specific compliance regimes. Domain-specific expertise and experience can be critical in attracting investors, and will be important in markets, such as SE Asia, with big growth potential (in April 2021, Bentley Systems justified an acquisition for its Asian region-specific challenges and project dynamics).