01
cities that drive themselves
a zone instrumented once is a zone autonomous forever. a ring-road, a port, an industrial park — every vehicle that enters inherits level-5 capability without a single onboard upgrade.
why dots
a post-vehicle
future.
the car is not the answer. the environment is. instrument the space once, and every vehicle in it — taxi, truck, bus, buggy — inherits autonomy. no fleet-wide retrofit, no decade-long safety case, no per-unit $50k stack.
dots autonomous is the infrastructure layer of that future.
the vision
when roads think, vehicles follow.
02
any vehicle, any class
trucks, buggies, passenger cars, buses, dump trucks — the hard sensing and planning are done by the environment. the vehicle just needs a light v2x unit and a brake pedal to listen to.
03
autonomy as infrastructure
we treat self-driving the way the last century treated electricity — as a utility wired into the ground, not a luxury bolted to a vehicle. every public space becomes a compute surface.
04
mobility at ~70% lower cost
no lidar, no per-vehicle ai stack, no multi-year calibration. the economics collapse, and autonomy reaches the 95% of vehicles that av 1.0 could never afford to touch.
the beliefs
three things we hold to be obvious.
- 01
the vehicle is the wrong place to be smart.
a car sees its own bumper. a city sees the whole street. perception scales with vantage point, not horsepower — so we moved the cameras off the hood and onto the pole.
- 02
one stack, every class.
passenger, freight, transit, industrial — they all share the same roads, but av 1.0 built a separate autonomy stack for each one. we don't. noiva is oem-agnostic, vehicle-agnostic, and class-agnostic by design.
- 03
cost is the only safety case that matters.
autonomy that's technically level-5 but economically level-0 helps no one. the moment a city can buy a mile of autonomous road for less than a fleet of lidar-equipped taxis, the future tips.
the shift
from vehicle-led to infrastructure-led.
av 1.0
every vehicle learns to drive
- · $50k+ of lidar, radar and compute per vehicle
- · years of per-oem calibration and safety cases
- · a separate stack for every oem, every model
- · performance degrades in edge conditions
- · cost scales linearly with every unit
av 2.0 — the dots way
the space learns. vehicles follow.
- · one instrumented zone, unlimited vehicles
- · light onboard unit — weeks, not years, to deploy
- · drift-free perception from a fixed vantage
- · oem-agnostic; every class rides the same grid
- · cost amortises across the fleet
the founder
Bhargava Hegde
founder & ceo · bengaluru · india
“if the last century taught the car to drive, the next one teaches the road.”
Bhargava started dots after a decade spent at the edges of autonomy, v2x and smart-infrastructure — long enough to see the same pattern repeat: every new av program paying to re-invent perception on a vehicle that, by design, can't see past its own bumper.
He founded dots autonomous to move the problem off the vehicle entirely. noiva — the stack that powers dots — is a clean-sheet answer: a single camera-led pipeline that lives in the environment, and broadcasts intent to any vehicle that listens.
His conviction is that autonomy becomes inevitable the moment it stops being expensive. the bet is that infrastructure, not per-car compute, is what gets it there.
build the grid. inherit the future.
cities, ports and campuses partner with dots to become the first places where every vehicle drives itself — by default.
start a pilot →