Today, companies at the vanguard of AI (Artificial Intelligence), are using the power of video intelligence analytics to exploit the massive volume of data generated by the technology. Vikram Gupta, founder, Awiros, was also drawn to this area after the completion of his Ph.D. at Carnegie Mellon University. Initially, he founded a company focused on providing advanced analytics solutions for the retail sector. Over a period of time, his experience with working with different types of IoT sensors helped him zero in on camera-based advanced analytics as the key technology that would shape the next leap forward in the digital age.
Realising the critical role a camera plays in sensor technology, Vikram, along with a team of developers, attempted creating video analytics applications for the retail sector. Soon, they discovered the multiple use cases of the technology, including security, facial recognition, surveillance and smart cities. However, despite the immense potential of video analytics to solve real life problems, the company was unable to manage such a large number of applications. This is because the technology is in inherently riddled with prickly issues.
First, video intelligence solutions are siloed and don’t allow communication between distributed systems. Second, such solutions are cost intensive in terms of scale and hardware requirements. Third, these need to be reconfigured for each use case scenario. And fourth, video analytics solutions are both time consuming and require a resource-intensive process.
Further, the average development time ranges from a few weeks to several months. The model also requires people of varying skill sets, which could prove to be an uphill task for an organisation. Critically, even once the right solutions are discovered, developers and modest startups find it difficult to reach out to prospective customers. Finally, the technology gives rise to compatibility issues as every new solution must be integrated with every other component of the surveillance infrastructure.
Such challenges led Vikram and his team to build an operating system — the first-of-its-kind for video intelligence. “The scope of video AI and computer vision is so vast that all the solutions cannot be catered to by a single company,” says Vikram. “It requires a community of developers.” For this purpose, a solution marketplace was created that was integrated on the operating system. And thus, Awiros was born.
A revolutionary idea in video intelligence
Awiros was conceptualised in 2018 as an OS that is capable of running a variety of video analytics applications. The idea also resonated with the founder’s doctoral thesis, which was about building an OS for managing multiple IoT devices.
In its early days, the company was focused on working with embedded devices. At this juncture, it was named “Awidit Systems Pvt. Ltd” for a logical reason. The Sanskrit word “Awidit” means something that is not immediately apparent, which is the inherent nature of any embedded system. Once the company created the unified OS for video analytics applications, the brand name “AwirOS” was coined — a clear messaging strategy to its targeted audience.
A conduit for enhanced digital transformation
Starting out as a bootstrapped venture initially, the startup received seed funding in 2018. Vikram and his co-founder, Yatin Kavishwar, saw the opportunity that AI-enabled video analytics presented in terms of effectively capturing data. “Video is the most pervasive form of data, forming close to 82 per cent of the data on the internet and CCTV cameras alone produce close to 800 PB of video data every day across the globe,” he says. “The sheer abundance of video makes it impossible to analyse all this data manually.”
It has been seen that AI-enabled video analysis alone can highlight data points that may be impossible to identify otherwise. The other advantage of video technology lies in its ability to lend itself to further analysis and extrapolation of information in multiple ways.
Vikram elaborates on the business model, “Initially, we created many applications on the platform to expand our customer base. We believed this would incentivise developers to develop applications on our platform.” The startup continues to strive to onboard developers from across the globe to cater to a wide variety of use cases.
Currently, the Awiros OS platform holds around forty solutions for a wide range of industries and is open to AI and computer vision developers across the globe. While developers can contribute their solutions to the app marketplace, consumers are able to search for unique use case solutions on the platform. To illustrate this, developers on the app marketplace have developed an application that detects whether workers in manufacturing and construction industries are wearing proper protective equipment while working on the field.
Another use case scenario that holds relevance in today’s times is how video intelligence is being leveraged to help stem the Coronavirus pandemic. Awiros is exploring partnership opportunities with MSIs (Master System Integrators), to identify Covid-19 patients in hospitals through facial recognition technology and track their movement within a hospital premises. This is important as such people need to be quarantined and prevented from mingling with the general public.
A bumpy ride
As Awiros sought to incorporate video AI in an already existing camera-enabled surveillance ecosystem, the founders strategically reached out to all the leading players in this space. The startup began pitching its product to hardware manufacturers such as Dell, Cisco and Intel; video management systems and command and control system developers such as Milestone, Tyco systems and Honeywell; and MSIs such as L&T and HFCL. The company also targeted — through extensive partner collaborations — potential customers who could use video AI in their business applications.
However, the entrepreneurial path has not been without road bumps for Vikram. He reveals that while customers seem to be excited by the variety of solutions and the ease of deployment that the OS offers, getting developers on board to develop applications on the platform is more of an uphill task. He is of the view that only when the platform gains more traction and recognition will it lead to more developers getting onboarded.
A savvy startup
Awiros is a part of the third cohort of T-Hub and acknowledges the value addition through the association. According to Vikram, T-Hub has opened doors for vital networking and mentoring opportunities. “As a result of the UTC Innovation Challenge, we are also getting to work with Pratt and Whitney, one of the biggest manufacturers of aircraft engines in the world,” he says.
Further, Awiros prides itself in being active on social media platforms such as LinkedIn and Twitter. It realises how critical it is to connect with both existing and potential business partners and customers to keep them updated on the company’s latest developments.
Besides social media channels, the company website also publishes regular blogs on the latest developments in the field of Computer Vision, AI, and the startup’s experiences in the development process. The founders aim to leverage their website to attract more developers to the Awiros platform. “We hope the Awiros community will continue to grow and become a major forum in the field of computer vision and AI in the times to come,” says Paurush Dube, the company’s product manager. He adds that events such as online hackathons are in the pipeline to generate interest in the Awiros OS as a development platform. “The idea is to attract the best of talents to the platform, and in turn, deliver the most advanced applications to our customers,” he says.
Smart solutions for a connected world
Vikram recognises the fluid environment in which the company operates. “An important learning has been to never underestimate the speed at which technology advances,” he says. “One always has to keep pace with the most recent technological advancements in order to stay competitive. Other than that, we are constantly working to improve the integration ability of the platform and make it as user-friendly as possible for consumers and developers.”
The company currently has around 25 clients, including government and private organisations. But its larger objective is to become the default development and delivery platform for video AI applications. Awiros has already partnered with several smart city projects across the country and hopes to be associated with more in the future. “This would help us in expanding our customer base, which would in turn help us in attracting more developers onto the platform,’ says Vikram.
For instance, Awiros has been associated with the Jaipur Smart City project. It partnered with Cisco that was the MSI for the project. In doing this, Awiros showcased its video intelligence capabilities through a use case for vehicle counting and classification across the Jaipur city.
Awiros has also been a part of Pune Smart City’s Center of Excellence (CoE) that strives to nurture upcoming technologies, such as video AI. While Awiros provided the software capabilities, Dell provided all the hardware requirements for a largescale implementation of the Smart City program. Awiros has been an integral part of PUDX (Pune Urban Data Exchange) and analysed video data for the Smart City program. The company leveraged its OS to analyse videos for several use cases. It resulted in the generation of meta data for further analysis by the Pune Smart City team. Subsequently. the technology generated valuable insights. For instance, a problem in the municipal department may be solved by analysing data from the police department.
The company is also working on the outreach of its developer partner program “Spira” to build confidence in the Awiros OS as the “go-to” development platform amongst AI and Computer Vision researchers across the globe. Through “Spira”, Awiros approaches developers to onboard their solutions on the platform with the assurance to expedite the development of any kind of application by ten times.
“Spira” offers developers a Software Development Kit (SDK) to convert their crude algorithm into a deployable product on the app stack within fifty lines of code or less. The initiative saves developers time, effort and resources. Spira is a channel to monetise developers’ applications when they create an algorithm and to resolve various issues that they might encounter. Their solution is showcased as an individual containerised application on the platform and leverages the OS capabilities solely for drawing and processing videos. Crucially, developers are assured complete intellectual property ownership of the code and solution. Awiros also offers a comprehensive revenue sharing model to its developer community that offers the developers a bigger percentage in the revenue their application generates.
Tackling new challenges posed by Covid-19
The unprecedented challenges posed by the pandemic has changed the paradigm of work for this startup. Awiros, which didn’t have a work from home policy until recently, had to embrace the new work culture, like many other organisations around the world.
The larger team has been able to collaborate through daily virtual ‘all hands’ meetings, which are held twice a day to discuss individual targets and deliverables. While this has been good for boosting the morale of team members, Vikram believes the new way of working has delayed decision-making processes that are usually swifter when employees are physically present in the office.
However, the team is trying to ramp up efforts in the current circumstances. In the absence of on-site client meetings, Awiros’ employees are largely focused on development, marketing and outreach efforts. The company has increased its social media activities and set high targets for individual deliverables during India’s 21-day lockdown period. Besides, it is also using the additional time on hand to incorporate cutting-edge technologies into the company’s products.
Vikram adds that the startup’s customers, such as the governments with whom it has partnered for Smart City projects, have been understanding of delayed timelines. Although Awiros is ready with its clients’ products, the startup is unable to deploy it at this point. The team is utilising the lockdown period to show demo versions to clients and seek their feedback on further iterations, if required.
“There will be losses and setbacks that every startup all over the world will be facing now,” Vikram says. He concludes, “The only thing we can do at this point is to minimise it as much as we can. We are hoping to come out really strong on the other side, whenever the Coronavirus crisis ends.”
We at T-Hub help startups gain access to better technology, talent, business networks and funding. Be a part of T-Hub Community. Join us here.