99x, a global technology services leader, continues to advance the frontier of innovation through its specialised 99x Yantra Labs team, focusing on the Internet of Things (IoT), embedded systems, and sensor-based applications. The team’s latest developments include use cases such as sensors that guide optimal solar panel placement, extending CCTV feeds to detect fire hazards, and several other bespoke customer engagements.
Commenting on the team’s line of work, Sachith Perera, Chief Technology Officer – Product Engineering at 99x, stated, “We have seen a strong growth trend among customers who offer solutions integrated with IoT sensors or embedded hardware devices. Some of these platforms require sub-millisecond response times that cannot be achieved using software alone. This requires embedding the decision logic into the hardware and building custom components to enable these response times.”
Dilshan Jayakody, Technical Architect, heads the Yantra Labs unit at 99x. Having over 20 years of experience in firmware development and low-level driver design, his automatic fluid level controller was honoured by Microchip Technologies Magazine as its featured project in September 2025. In October, Yantra Labs also hosted a customer visit from Norwegian Subsea, where Tonja Joseph, an Embedded Software Engineer, was in Sri Lanka for two weeks to collaborate with the embedded systems and IoT teams.
Reflecting on recent innovations, Dilshan shared, “Conventional fire detection systems rely primarily on smoke or heat sensors mounted on ceilings. While widely adopted, these systems are inherently reactive as they only trigger after smoke or heat has reached the sensors.
“Our innovation addresses this gap by leveraging computer vision and machine learning to detect visual cues of fire or smoke before traditional sensors respond. Processing camera feeds on the Jetson Orin NX at the edge enables real-time detection without relying on cloud latency.
“We have also built a solution to address the impact of shading on solar panel installations. When there is partial or severe shading on one panel, it creates a bottleneck that reduces overall performance across connected modules, resulting in a 10–30% energy loss depending on local conditions.
“Our Solar Shading Analyser solves these issues through a data-driven, site-specific approach that optimises layouts using solar data loggers equipped with GPS modules, pyranometers, and LoRa (long-range) mesh networks to collect and analyse solar intensity data, modelling shading patterns to recommend the best panel layout.”