Physical AI for Autonomous Mobile Manipulation
Robots shouldn't require the environment to be prepared. They should prepare themselves.
Okibo has developed an autonomous mobile manipulation platform that enables robots to enter unknown indoor environments and immediately begin performing industrial work – without digital twins, markers, pre-scanning, offline modeling or robot programming.
Instead of preparing the environment for the robot, the robot prepares itself for the environment.
The Fundamental Difference
Traditional robotic systems depend on engineering the environment before work can begin.
Conventional Autonomous Mobile Manipulation workflow
- Capture the environment (3D scan and/or BIM/CAD)
- Build or prepare a digital model
- (Often) install markers or reference points
- Generate and validate a task and motion plan (typically offline or in the cloud)
- Download the execution plan to the robot
- Deploy and execute
Many commercial solutions simplify the problem by relying on a human operator for navigation, positioning, or task setup.
Okibo's Autonomy Platform
- Deploy the robot
- Start working
The robot autonomously understands the environment, localizes itself, generates the work plan, plans its motion, avoids collisions, and executes the task , in real time, directly from the live environment.
How It Works
Our autonomy stack continuously performs every step required for autonomous mobile manipulation while the robot is working.
Understand
The robot continuously builds a live virtual model of the environment.
Localize
Virtual-model localization engine aligns the robot to its evolving world model, with 5 mm 6-DoF positioning accuracy.
Plan
Robot automatically generates industrial-grade work plans, computes poses, and creates executable trajectories.
Execute
Manipulator motion, mobile platform motion, and collision avoidance are planned simultaneously and adapted in real time with precision.
Why Construction?
Construction is one of the world’s most challenging environments for robotics.
- Every building is different.
- The environment changes continuously.
- Nothing is pre-engineered for automation.
We chose construction because solving autonomy here requires solving nearly every hard problem in autonomous mobile manipulation. The result is a general-purpose autonomy platform that extends far beyond construction.
Beyond Construction
The autonomy stack is application independent.
Any robot that must autonomously understand an unknown indoor environment, generate work plans and perform precise physical interaction can leverage the same core technology.
The application changes.
The autonomy platform does not.
