Karnataka will roll out its locally developed KEO artificial-intelligence (AI) personal computer across government schools in the state, Chief Minister D.K. Shivakumar said on Friday, as the government seeks to expand access to AI-enabled computing beyond Bengaluru and into rural communities.
The state has begun with an initial deployment of 2,222 KEO units across Karnataka, with the programme targeting Arivu Kendras (libraries) and residential schools before expanding to government schools more broadly. The device, developed through the Karnataka State Electronics Development Corporation (KEONICS) and industry partners, is positioned as an affordable, open-source computing platform for students and other users.
The device is priced at Rs 18,999 and is designed as a compact computing system with AI processing at the edge. It supports connectivity options including 4G, Wi-Fi and Ethernet, while the system is built around an open-source RISC-V processor and Linux environment, with AI capabilities designed to operate locally on the device, allowing it to be used even in locations with unreliable internet connectivity. The rollout marks the next phase of a project first unveiled at the last year, with the state now moving from development and demonstrations to deployment.
From Rural Centres To Government Schools
The first deployment is being structured around locations where the devices can get sustained usage. According to Sharath Kumar Bache Gowda, chairman of KEONICS, the rollout covers 239 taluks, with KEO units being placed in selected Arivu Kendras. Another 31 residential schools, representing each district, will receive 10 devices each, along with training and handholding for users.
The broader objective is to move computing access beyond smartphones. At the launch, CM Shivakumar said the state wanted government-school students and teachers to have the opportunity to learn and work with emerging technologies. The government has described KEO as part of its effort to democratise AI access and prevent the next phase of the digital divide from becoming an AI divide.
Industry-Backed Development
The KEO project has been developed through collaboration between the Karnataka government, KEONICS and technology partners. At the launch, IT-BT Minister Priyank Kharge credited the OPLUS team for building the compact system and said the project was developed without government expenditure, with industry partners contributing the resources needed to build and scale it.
Bache Gowda, in an interview with BW Businessworld, said ExaNIM was among the companies involved at the chip-design level. He said the project uses the NB2 system-on-chip, described as a 64-bit quad-core Arm-based chip with about 8 GB of RAM, while KEO itself is being built as an open computing platform around the underlying hardware and software stack.
The project also has two formal deployment partnerships. Shikshana Foundation signed memoranda of understanding (MoUs) with the Karnataka Residential Educational Institutions Society and RSA at the launch, according to the event proceedings.
Open-Source Model Keeps IP Unlocked
KEO’s intellectual-property model is deliberately different from a proprietary computing platform. Bache Gowda told BW Businessworld that the system has “no protected IP”, with the hardware and software stack built around open-source principles. “We don’t want to have a walled garden where I am protecting what information I have,” he said, adding that the platform was intended to remain open at the hardware, software, operating-system and application levels.
The KEO platform uses RISC-V architecture and an open-source Linux operating system. Bache Gowda said the approach would allow developers to build modular chip designs and applications for different requirements rather than being locked into a proprietary ecosystem.
That approach also leaves room for KEO to be adapted for applications beyond education. Bache Gowda said the underlying chip can support object and image recognition and could be used in applications such as industrial sensing. He cited a potential use case discussed with ONGC for monitoring drivers for signs of drowsiness.
Goa, More States On Roadmap
The state is also looking beyond Karnataka. Bache Gowda said the Goa government had expressed interest in deploying KEO in its schools and education system after the device was presented at an AI conclave in Delhi. “Goa wanted this for their schools, their education system,” he said. KEONICS expects KEO to find applications in other states and sectors as the platform scales, he added.
KEONICS is targeting a significant increase in deployment over the coming months. Bache Gowda said around 3,000 additional orders were already in the pipeline and that the combined deployment could reach about 5,000 devices within three to six months, with an eventual ambition of 10,000 to 15,000 units.
Bache Gowda added that revenue was not the primary objective and that the project was intended to make computing available to a wider section of the population. The state’s deployment strategy therefore places the first wave of devices in public learning centres and schools, with the longer-term plan extending KEO into other states and use cases.
Leave a Reply