A dancing robot still needs a person to decide
On 13 November 2026 twelve researchers published a proposal for what they call agentic robotics: a robot built in three layers, a way to test the whole system, and four levels of autonomy. We placed Paqarina inside it. Our robot has the two lower layers, and the layer that decides is still a person with a remote control.
The proposal in a few lines
The paper is called Envisioning Agentic Robotics: A New Era of Robotic Systems.
It starts from a difference. In the older picture a robot is one learned policy: an instruction goes in and movement comes out. In the new picture a robot is an agent: it pursues a goal by choosing among skills and tools, checking what happened, and changing its decision.
Then it describes three layers.
| Layer | What it does | What the paper puts there |
|---|---|---|
| System 2 | Decides. It reads what the person wants, chooses skills and tools, checks the result and changes the plan | A large model, and around it a harness: a planner, a memory, a verifier, an interface to tools and a manager of context |
| System 1 | Holds the skills that can be called | Learned movement skills, perception, motion planning |
| System 0 | Moves the body in real time | Whole-body tracking, and the control of each joint |
In the paper the word harness means the software that surrounds the model. It is not the strap that holds a robot up.
Beside the three layers runs a small monitor. It watches the person, the body and the scene, and it can stop the robot.
Where Paqarina sits
| Layer | What Paqarina has | Where it runs |
|---|---|---|
| System 0 | The network that follows a dance and keeps the balance: 29 angles, 50 times a second. Under it, the motors' own control | On the robot |
| System 1 | Eight dances. Each one is a skill, started from a remote control | On the robot |
| System 2 | A person, who chooses the dance, starts it, watches, and decides what to do after a fall | Beside the robot |
| The monitor | A person standing next to the robot | Beside the robot |
The paper says something close about the robots that work today: where a robot is trusted with more, a person in the background is doing the work of System 2.

The parts of a System 2 that already exist
Paqarina has built parts of what the paper calls the harness. They work, and none of them runs on the robot. They run on a laptop, days before.
| Part | What Paqarina has | When it runs |
|---|---|---|
| Memory | The archive: 137 takes of our own, kept as 26 points for every instant | Before |
| Planner | A score in words becomes a remix that respects a stage and a partner | Before |
| Verifier | Every dance runs 20 times in a second simulator. A planned sequence is checked against the robot's joints | Before |
| Interface to tools | The chain itself, from four cameras to 29 angles. A person runs each step | Before |
| Safety gate | A hoist and a person | During |
So the distance between Paqarina and an agentic robot is not that the parts are missing. It is where they are and when they run. Ours decide before the dance. An agent decides during it.
Four levels, measured on a dance
| Level | In the paper | For a robot that dances | Paqarina today |
|---|---|---|---|
| 1 | Carry out one skill | Dance one dance from beginning to end | In simulation, 8 dances of 8. On the real robot, 6 seconds of one |
| 2 | Complete a whole task without help | Receive a score and a stage, and dance it: choose the pieces, stay inside the circle, begin and end standing | The choosing is done ahead of time, on a laptop. The robot does not choose |
| 3 | Coordinate steps that depend on each other | Hold a place in a piece with dancers and avatars: an entrance, a partner at a distance, the music as the clock | Written by hand in the app, and planned as a duo. Not on a robot |
| 4 | Hold a role over a long time | Be a member of a company: rehearse, learn a new dance, remember a stage | Not started |
On this ladder the simulated robot stands on the first step, and the real one is below it.
Test the system, not only the model
For us the strongest point of the paper is about testing. It says that how well a model decides, and how well the whole system works with the world running, have to be measured separately. It also says the result should be reported as a profile across nine dimensions, not as one number.
Our record of November is a small case of exactly that. Every measure of the model was good: eight networks, each standing 20 times of 20. The system lasted six seconds. The cause was not in the network. It was in the first two seconds of the dance the network had been given, a pose with locked knees. It sat between two layers, where a test of one layer does not look.
This is Paqarina on the paper's nine dimensions. The fit is loose in places: they were written for robots that fetch and tidy, not for robots that dance.
| Dimension | What we have measured | What is missing |
|---|---|---|
| Knowledge | How close a novel suri, written by the model, is to real suri: 8.1 cm per point | Dances it has never seen |
| Understanding | The robot reads 154 numbers about its own joints, its tilt and the dance | It does not see the stage, the partner or the person |
| Reasoning | A plan for a score: 15 pieces of 15 in the dance asked, and 1.69 m reached where 1.5 was asked | Changing the plan after a fall |
| Tool use | The tools of the chain are run by people | The robot calls none |
| Memory | 137 takes of our own, 4.4 hours | The robot remembers nothing from one run to the next |
| Cross-embodiment | The same 26 points on an avatar and on the robot, fitted within 2.7 cm. The ankle at its limit 24 per cent of the time | A second robot |
| Safety | Falls counted in simulation. On the floor, a hoist and a person | A robot that notices the risk itself |
| Efficiency and cost | More than 250 dollars in all. More than nine hours of practice for each dance. Four minutes for the planner | The cost of a decision made during the dance |
| User intent | The choreographer's score, in words | The dancers' judgement of what comes out |
What dance adds to the proposal
Three things matter for dance and are thin in the paper.
The first is time. The paper notes that System 2 takes seconds to decide while control runs in thousandths of a second. A dance step lasts less than a decision. For dance the deciding has to happen between two phrases, or before the music starts. That is why our plan is made ahead.
The second is the body in motion. The paper names dynamic control and walking as open problems. Dance lives there: one foot in the air, the hips turning, the weight never still.
The third is whose dance it is. The nearest dimension is user intent, and it assumes one user. A dance of a community has several: the choreographer, the dancers, and the community the dance belongs to.
What we ask the robot to do
By layer, this is what we ask:
| Layer | What we ask | Size |
|---|---|---|
| System 0 | Hold 29 angles, 50 times a second, and stay on its feet | From 31,088 to 46,313 angles for one dance |
| System 1 | One dance at a time | 8 dances, from 21 to 32 seconds |
| System 2 | Nothing. A person decides | 0 decisions made by the robot |
Thanks
Paqarina is a Mozilla Fellowship project.
Thanks to Dr Pan Hui, Professor Xin Tong and Jie Yu for allowing us to test our policies on the robot.
What the record shows
| Measure | Value |
|---|---|
| The paper | Envisioning Agentic Robotics: A New Era of Robotic Systems, 13 November 2026, 12 authors |
| Layers of the proposal that run on our robot | 2 of 3 |
| Skills on the robot | 8 dances |
| Decisions the robot makes | 0 |
| Parts of a System 2 that Paqarina has, off the robot | Memory, planner, verifier |
| Level reached in simulation | 1: one whole dance, 8 of 8 |
| On the real robot | 6 seconds of one dance |
| Evaluation dimensions in the paper | 9 |
What remains open
Whether the verifier can move from the laptop to the moment before the dance: is the robot standing, and is the first pose one it can take. Whether a decision can be made between two phrases without missing the beat. Whether the dancers' judgement can be written as one of the dimensions. The layer that decides is still a person with a remote control. The first thing that person could hand over is a check: whether the robot is standing before the dance goes in.