An Automotive Analytics Engine made by fusing a Rule-Based Expert System with a ML layer – A Domain Aware AI.
EVA® was conceived with one vision – value providing analytics. All engineering elements like form-factor, connectivity etc. were fundamentally part of the design for impactful results, not because they were trends & buzz words.
EVA® has been integrated throughout the automotive ecosystem.
EVA® Deployment Architecture
EDGE
EVA's platform-agnostic analytics library/SDK is designed for seamless deployment on edge devices, telematics units, IoT platforms and smartphones, delivering ultra-low-latency analytics for real-time safety, performance and reliability.
CLOUD NATIVE
EVA integrates seamlessly with existing mobility platforms through standard APIs and data pipelines, enriching them with deep vehicle analytics while preserving existing workflows.
HYBRID
EVA's hybrid architecture combines edge-based real-time analytics with cloud-scale intelligence, delivering immediate operational decisions and long-term engineering insight throughout the vehicle lifecycle.
Observational AI (OAI)
An in-vehicle AI layer that continuously observes vehicle operation, driver behaviour and the surrounding environment to develop a contextual understanding of vehicle operation in real time. Powered by EVA®, OAI performs situational assessment, behavioural reasoning and recommendation logic, enabling proactive guidance that is delivered through its complementary operational manifestations— PI-ADAS and HVI.
“OAI was originally conceived to enable every driver and rider to realize the full engineering intent of their machine through effective vehicle energy management and responsible vehicle operation. It was never designed to replace the driver—it was designed to help people become better custodians of their machines. What began as solutions to individual engineering challenges gradually revealed a common thread. As these capabilities evolved, it became evident that they were manifestations of a broader intelligence. That realization ultimately led to what is today known as Observational AI.”
Powertrain Informational ADAS (PI-ADAS)
The Powertrain Informational ADAS (PI-ADAS) is OAI's operational manifestation for optimum motive power control. By continuously observing powertrain behaviour, it delivers proactive, real-time guidance on throttle and transmission control for effective vehicle energy management, reducing tailpipe emissions, enhancing safety and extending powertrain durability—without directly controlling the vehicle.
Human Vehicle Interaction (HVI)
The Human Vehicle Interaction (HVI) layer is OAI's operational manifestation for optimum vehicle operation. By continuously observing vehicle operation and driver behaviour, it delivers proactive, real-time guidance on responsible vehicle practices, intelligent use of vehicle features and contextual operational assistance. HVI observes and guides the driver/rider to improve safety, enhance vehicle efficiency and promote responsible vehicle operation.
Predictive Maintenance
A deterioration assessment layer that uses CAN signals, degradation models and proprietary Powertrain Physics Models to quantify system deterioration, predict the remaining useful life of components and prescribe maintenance before functional failure occurs. To achieve broad vehicle coverage, EngineCAL has specially engineered models that operate effectively with standard OBD-II data across passenger cars, commercial vehicles, two-wheelers and other vehicle segments. This enables scalable deployment, particularly across the aftermarket, including legislation compliance initiatives.
Vehicle Intelligence
A Deep Vehicle Analytics layer that transforms raw vehicle data into actionable engineering intelligence across the entire vehicle lifecycle. By continuously evaluating vehicle performance, operational behaviour and system interactions, it enables objective assessment for OEM R&D, End-of-Line quality, legislation compliance and aftermarket applications, improving vehicle energy efficiency, durability and overall engineering effectiveness.
