How designing autonomous robots for real South Asian environments changes imported engineering assumptions.
Why lab-proven autonomy fails in the field
Most autonomous robot research assumes flat floors, predictable lighting, and reliable Wi-Fi. South Asian field conditions present a harder problem: power cuts, dust, humidity, cable clutter, and human traffic patterns that no simulation captures.
This post documents what we learned designing the ARIOT floor-cleaning robot prototype for real environments — and why every assumption we imported from Western research labs needed re-examination before we shipped anything.
SLAM on real surfaces
Our LiDAR-based SLAM pipeline performs well on paper. On a typical Dhaka office floor — worn tiles, scattered furniture, extension cables, and partially-blocked corridors — it requires significant tuning that no off-the-shelf algorithm provides.
Key failure modes we hit:
- Reflective surfaces — polished tiles and glass partitions produced phantom obstacles in the LiDAR point cloud, causing the map to distort on the second loop.
- Cable clutter — charging cables, extension leads, and cable trays created false-floor readings that triggered the cliff sensor incorrectly.
- Moving obstacles — people walking slowly through the mapped zone caused the local planner to re-plan constantly, producing oscillation near doorways.
The fixes for each of these — sensor fusion weighting, obstacle inflation radius, and recovery behaviour — are documented in our internal test logs and will be published here once validation is complete.
Power management for a 220V country with brownouts
Battery management for a robot that docks and recharges in a market with frequent voltage sags is not covered in most embedded robotics textbooks.
The design constraints:
- Input voltage must handle the wide voltage swings common on Bangladesh commercial grid connections.
- The charging circuit must recover cleanly from a mid-charge brownout without leaving the cell in a partial-charge state.
- The robot must reach the dock when battery drops below a threshold — and that threshold must account for the distance to the dock, not just a fixed percentage.
The charging circuit design, brownout recovery behaviour, and dock-range logic are documented internally and will be published with the product datasheet once approved.
Thermal performance in a hot, humid climate
Bangladesh summers push ambient temperatures well above the operating range tested in European laboratory environments. Components that run within spec at 25°C may derate significantly at 35–40°C ambient with 80%+ relative humidity.
Thermal derating data, heatsink sizing, and enclosure ventilation are under engineering review; published figures will follow validation.
What we learned so far
Designing for South Asia means treating the environment as a first-class design constraint — not an afterthought applied after the algorithm is "working".
Every field test reveals something a simulation did not predict. We document each failure honestly here because the engineering community in this region benefits more from honest failure reports than from polished success stories.
This post is a living document. Specific figures, failures, and fixes will be added to this post as field validation progresses and each result is approved for publication.
ARIOT Engineering Team
[Author bio pending — will appear once engineering team profiles are approved.]