The analytical fix

Estimate the payload before anything learns it.

Augment an EKF with the payload mass and one disturbance torque per joint, measure what the servos report, and feed the estimates back as goal offsets. The result is the baseline every learned method must beat.

Written lesson + build labEvidence-based gate

The goal for this build

What you’ll be able to do.

Derive the Kalman update and build an EKF with payload mass and per-joint disturbance torques in its state.
Explain what a still hold cannot observe and design an excitation move that can.
Feed bounded estimates back as goal offsets and rerun bench.py.
Judge an estimate by its error, its 2σ and its innovations, and say where current resolution limits it.

What to have ready

Module 2 controllers and their bench.py results table
Gaussian distributions, mean and covariance
Matrix products, transposes and solving linear systems
The Module 1 observation model: register resolution, delay and noise

Download the course starter ↓

Free account

Read the lesson, run the lab, pass the gate.

The whole course is free. Create an account to unlock the lesson, the build lab, the task and the gate command for this module, and to download the starter kit.

Create a free account Sign in

03 / Prove it

Keep the evidence.
Then run the gate.

What you ship

estimators/ekf.py with documented states, noise models and checked Jacobians
Still-hold and excited-hold estimate plots from priors at least 50 g off
A bench.py rerun with EKF feedback beside the Module 2 controllers, on the same seeds
The Weigh it result in sim, with its 2σ
A Module 3 line in PROGRESS.md

Module 03 gateEvidence required

From a prior at least 50 g off, the payload estimate stays within 10 g of the true mass from 2 s into a hold plus a small excitation move.
Over the same window, its 2σ stays within 15 g.
On the +150 g case, tracking error after those 2 s stays within 1.5× nominal or nominal plus 1 mm, whichever is larger.
There is no overload trip on the +150 g case.

With a free account: the gate command and how to record your evidence.

The story behind the build

Episode companions.

EP 03Teaching a robot arm to weigh things

Written companions are available now. Recordings and customer access are forthcoming.

Made it through the lesson?

Save your reading progress here. Record your gate result in your project.