Some robot vacuums have a better sense of direction than others

Say hello to our robot vacuum test room. Inside it are objects that might look a little odd, but they're designed to simulate furniture and obstacles a robot will encounter in the real world.

We watch what robot vacuums do inside the test room closely.

A camera captures robot vacuum activity from the ceiling.

Shown here is a structure that mimics table and chair legs. 

Robot vacuums interact with these objects as if they're real furniture.

This box imitates the outline of a living room cabinet or side table. 

To break in our new test room, we ran nine robot vacuum models across its floor.

Basic robot vacuums equipped with merely collision and proximity sensors, such as the Eufy RoboVac 11S Max, bounced around our test room randomly without covering the entire floor space. 

The Ecovacs Deebot 500 is another robot vacuum with basic, random navigation. You can see here it covered our test floors inconsistently. These vacuums also tend to run for a long time.

The iRobot Roomba i7 Plus carved out this pattern while cleaning our test room floor using optical navigation.

Robot vacuums with optical sensors and real time mapping tend to be faster and cover floors better. This is the coverage are we logged for the iRobot Roomba i7 Plus.

Equipped with a LIDAR laser navigation and mapping system, the Neato Botvac D6 Connected followed a precise path through our robot vacuum test room.

The Neato Botvac D6 Connected robot vacuum used its laser LIDAR system and SLAM algorithm to good effect. It offered swift clean times, and plenty of floor coverage.

The Electrolux Pure i9 vacuum uses a hybrid navigation and sensor system. Even so its path definitely missed areas of our test room floor.

The Electrolux Pure i9 coverage of our test room had plenty of gaps.