MODULE 01 / FOUNDATIONS
Meet the robot
Build your first map of BDX-R: what holds it together, what moves it, what senses movement, and what makes decisions. Start with observation before installing software.
By the end, you can…
- Name what the robot’s structure, actuators, sensors and computer do.
- Explain how a measurement can change the robot’s next movement.
- Explain what a 3D viewer shows and what still needs to be tested.
None. A browser and a notebook are enough.
Meet the systems that make a robot move
Imagine BDX-R lifting one foot. Its structure holds the pieces together. An actuator moves a joint. Sensors measure what happened, and a computer uses those measurements to decide what to request next. A battery supplies the energy. The robot needs all of these jobs to work together.
An actuator is a device that produces movement or force using supplied energy. In this robot, an actuator includes an electric motor and other parts that help control its movement. A sensor has a different job: it measures something. For example, an encoder is a sensor that measures a joint’s position. Measuring a joint does not, by itself, move it.
- Structure: the frames, brackets and covers that hold or protect parts.
- Actuators: the devices that move joints.
- Sensors: the devices that measure joint positions and body movement.
- Computer and software: use information to decide what movement to request.
- Communication: carries commands and measurements between components.
- Power: supplies and distributes the electrical energy.
One foot, several jobs
To lift a foot, the structure carries the load, actuators move the leg, sensors report its movement, and software adjusts the requests. The battery powers the process.
Read the body from the robot’s viewpoint
Left and right belong to the robot. When it faces you, its left leg is on your right. A joint is a place where connected parts are allowed to move relative to one another. A knee is a familiar example: the lower leg moves relative to the upper leg.
In our selected BDX-R model, each leg has five joint movements: hip yaw, hip roll, hip pitch, knee and ankle. Yaw turns around an upright axis, like looking left and right. Roll tilts sideways, like leaning your head toward one shoulder. Pitch tilts forward or backward, like nodding. The labels describe the direction of movement.
A visible part and a joint are not the same thing. Several brackets can be fastened together and move as one piece. A head assembly can contain more than one joint. Use the model to locate possible movements; we will verify their exact axes in later lessons.
Measurements let a robot correct its movement
A target is what we want to happen. A measurement tells us what actually happened. Feedback means using the measurement to change the next request. You use feedback when you reach for a cup: you look at where your hand is and adjust it as you move.
For a robot joint, the process is: choose a target → request movement → measure the result → compare it with the target → adjust the next request. When the measurement comes back and influences the next request, we call this a closed feedback loop. “Closed” describes the returning information, not a closed wire circuit or a loop in a video.
Sending the same request again and again is not enough to make feedback. A prerecorded animation can repeat without checking the robot’s actual movement. Feedback needs a measurement that affects what happens next.
The knee has not reached its target yet
Suppose the target angle is 20°, but the joint-position sensor reports 15°. The software sees a 5° difference and can adjust its next request. The numbers illustrate the idea; they are not BDX-R operating settings.
What the computer does — and a word for later
A controller is software that turns a desired result and measurements into movement requests. You might ask the robot to move forward slowly; the control system must coordinate many joints to work toward that goal. The high-level request and the individual joint requests are different things.
Later in this course you will meet a policy: a rule, sometimes learned through training, that maps available information to an action. For now, think “a way of choosing the next action.” You do not need to understand policy outputs, training or model files to complete this module. We will teach those before asking you to interpret them.
What a 3D viewer can tell you
CAD means computer-aided design: software used to describe the shapes and arrangement of parts. A CAD viewer displays that design so you can look around it, zoom in and inspect represented shapes. Orbiting the camera means moving your viewpoint around the model; it does not mean the robot itself is moving.
A recorded animation plays a sequence of poses that was already chosen. A live physics simulation calculates how a modeled robot moves as time advances, using assumptions about things such as weight, forces and contact with the floor. A controller can use simulated measurements to choose new requests while that simulation runs.
An orbitable 3D picture therefore shows that you can inspect the represented shape. It does not establish that the robot balances or that its parts are ready to print. Balance needs control and physical testing; printing also needs decisions about dimensions, materials, orientation and fit. You will learn those checks in their own modules.
Look at the caption before deciding what you saw
If a view says “recorded replay,” the displayed walking was calculated or recorded earlier. Rotating the camera around that replay does not turn it into a new physics experiment.
Observe first, then explain
Write down what you can see before guessing why it happened. “The body turns while it moves forward” is an observation. “The feet are slipping” is an explanation that needs more evidence, such as foot motion while touching the ground. Keep an observation and a possible cause in separate columns.
BDX-R files and hardware change over time. Keep the source or view name and its revision, when available, beside your notes. Different models can include different joints. For this first lesson, learn the system roles and identify the body parts; later lessons will explain how to compare the exact software and hardware configurations.
WORDS YOU NOW KNOW
- Actuator
- A device that uses energy to produce movement or force.
- Sensor
- A device that measures something about the robot or its surroundings.
- Encoder
- A sensor that measures a joint’s position.
- Joint
- An allowed movement between connected parts.
- Feedback
- A measured result used to change the next request.
- Controller
- Software that uses desired results and measurements to request movement.
- CAD viewer
- Software for inspecting the represented shapes and arrangement of a design.
YOUR CALL
Finished with this module?
0 of 6 lab steps checked. 0 of 3 questions answered. You decide when the module is complete. Revisiting or editing it will not change your decision.
You can reopen it later. Checking every box will not automatically complete it.Go to the source
Primary references for this module. Use the version and hardware revision named in the lesson.
Kayden’s BDX-R build videoBDX-R project overviewBDX-R deployment source