Pangyo-based IT enterprise BI C&S has introduced AIBUS™, a conversational voice recognition AI platform designed specifically for outdoor public transportation infrastructure. Leveraging 27 years of data and AI expertise, the company aims to address the limitations of current touch-based and indoor-centric kiosks. By targeting smart city markets, BI C&S provides a solution for high-noise environments like bus stops and transfer centers. This move signals a strategic shift toward inclusive, hands-free public information systems that support visually impaired users, the elderly, and foreign travelers through advanced multilingual capabilities and robust acoustic engineering.
BI C&S Targets Smart City Infrastructure
BI C&S is positioning AIBUS™ to capture the B2G market by addressing accessibility gaps in public transit. Current kiosks often rely on touch interfaces, which create barriers for the visually impaired and elderly, or fail in noisy outdoor settings. To counter this, AIBUS™ utilizes a full-stack voice AI pipeline developed entirely in-house, covering acoustic pre-processing, Korean speech-to-text (STT), conversational LLM, and multilingual text-to-speech (TTS). This internal development avoids external API reliance, ensuring a rapid 1.5-second response time and enhanced data privacy for public institutions. The system supports 10 languages, including English, Chinese, and Japanese, and offers both voice-only and hybrid voice-and-touch modes. The company has already secured a 600 million KRW contract to supply 15 kiosks for a Ministry of SMEs and Startups project and signed an MOU with the Seoul Dongdaemun-gu Office for deployment in January 2026.
Triple Acoustic Pre-processing for High-Noise Environments
A primary technical challenge for outdoor voice AI is maintaining accuracy amidst ambient noise. AIBUS™ is engineered to achieve voice recognition accuracy exceeding 95%, even in roadside environments where noise levels reach 80 to 100 dB. The platform implements a specialized triple acoustic pre-processing technology consisting of beamwidth optimization, adaptive noise canceling, and bus stop noise pre-profiling. This technical cycle improves the signal-to-noise ratio (SNR) by approximately 10-15 dB, allowing the system to function effectively in high-traffic urban areas. Beyond technical specifications, BI C&S maintains a heavy R&D focus, with 62% of its workforce dedicated to research and development. Following domestic validation through 2027, the company plans to export its "K-Smart City Package" to Southeast Asia, the Middle East, Japan, North America, and Europe starting in 2028, targeting regions with aging populations or high multilingual demand.
Key Takeaways
- AIBUS™ achieves over 95% voice recognition accuracy in noise levels between 80 and 100 dB.
- The platform utilizes an in-house full-stack AI pipeline to ensure a 1.5-second response time.
- BI C&S secured a 600 million KRW contract to supply 15 AI kiosks for a government project.
TechInsyte's Take
In our view, BI C&S is making a calculated move to solve the "last mile" of smart city accessibility. By developing a full-stack pipeline rather than relying on third-party APIs, they have addressed two critical enterprise requirements: low latency and data sovereignty. This is particularly vital for government contracts where privacy is non-negotiable. The ability to maintain high accuracy in 100 dB environments suggests a level of hardware-software integration that generic AI models cannot match. This signals that the future of public infrastructure lies in specialized, environment-aware edge AI rather than general-purpose cloud solutions.
Source : https://www.einpresswire.com/
