How RoadVision AI Wants To Change Road Infra Management

There is hardly any sector or industry, from healthcare to finance, that has not been transformed by technology. However, despite its crucial role in economic growth, India’s infrastructure sector has seen little progress on the technological front.

From planning to building and maintaining roadways, manual methods continue to dominate every facet of the road development lifecycle in the country. This reliance on outdated practices not only slows things down but also leads to accidents.

The numbers tell a worrying story: In 2022 alone, India saw a shocking 4.6 Lakh+ road accidents, resulting in more than 1.6 Lakh deaths and over 4.4 Lakh injuries.

What’s more, these figures represent a significant increase compared to the previous year, with accidents rising by 11.9%, fatalities by 9.4%, and injuries by 15.3%.

To change the sorry state of the country’s road network by automating and optimising the road management process, Delhi NCR-based RoadVision AI is leveraging the power of new-gen tech, like artificial intelligence (AI) and digital twins.

Spearheading this cause is Hardik Dave, the founder of RoadVision AI, who incorporated the AI solutions entity early this year (2024) to transform road infrastructure management with digital twins and AI technologies.

However, this is not Dave’s first tryst with a startup. He took the entrepreneurial plunge after quitting his handsomely paying corporate job in 2020. Soon after stepping down from his role as the consultant of EY, he started an AI solutions company, Indika AI. According to the founder, the company is profitable, which allowed him to explore new avenues.

“We provide a road asset management system to companies like L&T, IRB Infra and others engaged in operations, maintenance of roads and other various aspects such as consulting, design, construction, and post-construction. Our AI system ensures that proper processes are followed for continuous operations and maintenance of roads,” Dave said.

A Sneak Peak Into RoadVision’s AI Tech Stack

RoadVision’s AI solution is a proprietary AI-powered Large Vision Model (LVM) that automates road survey data processing and analysis, replacing the manual visual road inspections by engineers, which are time-consuming, costly, inefficient, and prone to human errors.

Its LVM is trained on over 10,000 kilometres of road data to process the road images and videos to identify road conditions, classify them as per the type and size of road distresses, and calculate pavement distress ratings as per the industry and regulatory standards.

The solution also detects and catalogues lane markings, road signages, advertisements, and other road inventory and safety assets in alignment with the applicable guidelines.

At RoadVision AI, data analysis begins with the collection of visual and location data through a smartphone-based mobile application. This data is then processed and visualised using a user-friendly GIS-based web interface. Through the intuitive visualisation of digital twins and data-driven analytics, RoadVision enables collaborative monitoring throughout the road maintenance lifecycle.

For the uninitiated, a digital twin is a virtual representation of an object or system designed to accurately reflect a physical object.

While RoadVision AI offers a comprehensive solution for automated road operations and maintenance, it comprises three key components.

Its smartphone-based application streamlines data collection, gathering visual and location data effortlessly. This approach is particularly beneficial for companies where regular inspection via Network Survey Vehicles (NSV) is impractical due to high operating costs.

With the help of its proprietary AI engine, RoadVision AI automates the processing and analysis of road survey data collected through either the mobile application or any other method.

Moreover, its GIS-based web interface provides an intuitive visualisation of digital twins, supports data-driven analytics, and facilitates collaborative monitoring throughout the road maintenance lifecycle.

Dave claims that RoadVision AI’s solutions can save clients up to 90% in time and 80% in costs compared to their current road survey methods.

Initially, the startup focussed on smart cities due to their smaller road networks, unlike larger cities such as Bangalore.

“We then shifted our focus to highways and larger companies operating tens of thousands of kilometres of roads. We conducted a pilot project with L&T and engaged in discussions with the government. Additionally, we partnered with IRB Infra, India’s largest construction company, which oversees more than 15,000 kilometres of roads across the country, handling both construction and ongoing maintenance. We have similar partnerships in the pipeline,” Dave said.

RoadVision AI’s Business Model

RoadVision AI operates on a flexible business model that caters to both business-to-business (B2B) and business-to-business-to-government (B2B2G) sectors. It employs a freemium pricing strategy, offering free access to its mobile application and basic web platform features.

For clients requiring advanced functionalities like the AI model for automated data processing and premium web platform features, pricing varies based on factors such as the length of road to be surveyed and specific modules needed. This approach ensures flexibility and scalability for a diverse range of clients.

Despite being in the early revenue stage, RoadVision AI claims to have an order book exceeding INR 4 Cr for the current fiscal year. While bootstrapped startup’s primary markets are the US and India, the founder is receiving interest from regions like Australia, Southeast Asia, and Africa.

With one commercial deployment completed in Jalandhar and four more projects in the pipeline, RoadVision AI anticipates revenue of $500K (INR 4 Cr) for the financial year 2024-25 (FY25). These projects have completed the pilot phase and await renewal once the model code of conduct is lifted.

The Way Forward

The scope of AI in road monitoring is vast and promising. AI-driven solutions have the potential to transform road management by providing real-time insights, predictive analytics, and proactive maintenance strategies.

From detecting potholes and cracks to optimising traffic flow and enhancing safety measures, AI holds the promise of enhancing efficiency and effectiveness in road infrastructure management.

In a market brimming with road technology and survey companies, RoadVision AI faces competition from global players like RoadBotics and RoadAI, particularly in the realm of AI-powered road condition survey technology. However, RoadVision AI distinguishes itself by providing cutting-edge road management and infra-development solutions at an affordable cost.

RoadVision AI’s long-term aim is to use its AI-driven solutions to significantly reduce road accidents and fatalities caused by engineering deficiencies — the real-world problem that cofounder Dave takes pride in solving with its tech.

[Edited by Shishir Parasher]