This Delhi-based startup uses AI for smarter monitoring, video analytics

Believe it or not, a lot of firms are still reliant on dated tools for monitoring. For instance, CCTVs are pretty much common at factories, workplaces and other industries. You can easily visualise a big room with multiple monitors keeping a tab on things. And this has been around for years. The vast amount of this data generally goes unused and barely gives any leverage.

Delhi-based Wobot.ai is trying to resolve this big gap with ‘smarter video analytics solutions’ that leverages cutting edge technologies such as Artificial Intelligence to achieve more productive and automated monitoring. These solutions can be personalised and implemented across industries such as hospitality, manufacturing, and retail among others.

Wobot’s claim to fame has been its partnership with the Indian Railway Catering and Tourism Corporation (IRCTC) in 2018. The company helped the IRCTC leverage Artificial Intelligence to monitor the food production process across its several base kitchens. The monitoring tool helped detect anomalies as well as bring uniformity in passenger experience. Since then, the company has been working with many big names in the industry such as Kitopi, CureFit, Rebel Foods, Travel Food Services, Blue Tokai, and Apparel group.

ALSO READ: Why AI experts are trying to ban this neural model

The challenge with such CCTV-based monitoring solutions is achieving accuracy. The AI, however, has helped the firm achieve some success on this front. According to Wobot co-founder and CEO Adit Chhabra, the company’s secret sauce has been focusing on one use case and making sure that they have trained ample amounts of data, whether it is synthetic or augmented or just the feed from CCTVs.

“We focused on one industry and a few use cases within that industry. Eventually, we built an in-house training pipeline and annotation team to help train new models faster and continue to improve existing ones on a regular basis for accuracy,” he told Hindustan Times.

Under the hood is a cutting-edge neural network that is based on sophisticated models that are not resource-intensive and can deliver relevant monitoring.

“We amalgamate Activity and Object Recognition for our use cases. The analysed visual data is then delivered to relevant stakeholders through real-time notifications on the Wobot application, dashboards, or emails,” he added.

Chhabra believes industries other than Security and Retail will also embrace video analytics solutions. He also pointed out that the number of installation of CCTVs YoY is estimated to grow at a CAGR of around 11% globally. This will certainly help new and emerging companies to leverage video analytics for better output in the longer run.

ALSO READ: This AI tool can convert a blurry photo into high-quality portrait shot

He is also hopeful that the 5G network will transform the video surveillance segment. It is worth noting that the upcoming 5G is not just aimed at mobile networks but also at a variety of connected devices and appliances.

“The perpetual issues of bandwidth and latency can finally be solved by integrating 5G in the video and Data Analytics system. 5G allows us to send high-quality videos (1080p, 4K) over the network as it provides a high video transmission bandwidth. This will, in turn, reduce the latency associated with video transmission and video processing,” he explained.

“Another important aspect associated with 5G is its ability to integrate billions of IoT devices. The integration of IoT devices in the video and Data Analytics pipeline will provide us with massive amounts of data that we can leverage to draw actionable insights,” he concluded.