For years, the dream of the self-driving car has arrived in Indonesia as an expensive import, designed for the wide, predictable boulevards of Europe or North America. These machines rely on LiDAR, a sophisticated laser-based sensing technology that can cost tens of thousands of dollars per unit. But in the laboratories of Bandung, Widyotriatmo chose a different path, one dictated by the reality of the streets outside his window.
The innovation of the AVA (Autonomous Vehicle Adaptive) lies in its eyes. Rather than relying on laser pulses, it uses a network of adaptive cameras and computer vision AI. This system is trained specifically to recognize the sudden, weaving trajectories of the two-wheeled vehicles that dominate local traffic—a demographic often ignored by the training datasets of Silicon Valley.
To watch the vehicle navigate is to see a machine learning a specific cultural vocabulary. Where Western systems might stall in the face of informal public transport or roadside food carts, AVA's convolutional neural networks estimate depth and distance through optical lenses, processing the world with the same visual cues a human driver uses. It is a shift from brute-force sensing to intelligent perception.
The ambition of the project, a collaboration between ITB and local industrial partners like PT Sibernetika and PT TESA, extends beyond the university gates. Widyotriatmo sees these vehicles serving industrial estates and the terminal zones of airports. His motivation is a quiet, persistent form of national dignity: the belief that Indonesia should not merely be a market for the future, but the place where that future is built and understood.
As the Indonesian government prepares its new capital, Nusantara, to be a hub for autonomous transit, the work in Bandung provides a necessary foundation. It proves that the most sophisticated technology is often that which remains closest to the ground, reflecting the faces and the frenetic, living pulse of the place it is meant to serve.
The system is designed to match Indonesian infrastructure and road environments, not imported assumptions.