Network
Language · FA
Dark mode

ENGINEERING

Robotics

Robot types, actuators, sensors, kinematics, ROS 2 concepts.

Robot Types

TypeDegrees of FreedomUse CaseExample
Articulated arm4–7 DoFAssembly, welding, pick-and-placeKUKA KR6, ABB IRB 120
SCARA3–4 DoFFast horizontal assembly, dispensingEpson T6
Delta / Parallel3 DoFUltra-fast pick-and-place in food/pharmaABB FlexPicker
Cartesian / Gantry3–4 DoF (linear)Large workspace, heavy loads, 3D printingCNC routers, large printers
Collaborative (Cobot)6–7 DoFHuman-robot collaboration, safe contactUR5, Franka Emika
Mobile (AGV/AMR)Navigation DoFWarehouse logistics, last-mile deliveryBoston Dynamics Spot, MiR
Humanoid20–30+ DoFResearch, service, bipedal locomotionAtlas, Optimus (Tesla)
Soft robotContinuous deformationMedical, grasping delicate objectsFesto Bionic Cobot

Actuators

TypePrincipleTorque/ForceSpeedBack-drivable
DC brush motorElectromagneticLow–mediumHighYes
BLDC motor3-phase electromagneticMedium–highVery highYes
Stepper motorDiscrete stepsLow–mediumLowLimited
Servo (hobby)BLDC + gear + encoderMediumMediumPartial
Harmonic driveStrain wave gear reductionVery high (100:1+)LowLimited
HydraulicPressurized fluidVery highMediumNo
PneumaticCompressed airMediumVery highNo
Linear actuatorScrew/rack/solenoidMedium–highLow–mediumVaries
PiezoelectricCrystal deformationLow (nm steps)Very high frequencyNo

Sensors

SensorMeasuresOutputCommon Interface
Encoder (optical/magnetic)Angular / linear positionPulses, A/B/ZQuadrature, SPI
IMU (accel + gyro)Acceleration, angular rate, orientation6-axis dataI²C, SPI, UART
LiDAR3D point cloud (distance by laser TOF)Point cloudEthernet, USB
RGBD camera (RealSense, ZED)Color + depthImage + depth mapUSB3
Force/Torque sensor6-axis force and torqueAnalog / digitalCAN, EtherCAT
Tactile / pressureContact force distributionArray of pressuresSPI, I²C
Proximity (ultrasonic/IR)Distance to obstacleAnalog voltage, digitalGPIO, I²C
GPS / GNSSGlobal positionNMEA sentencesUART

Kinematics

ConceptDescription
Forward kinematics (FK)Given joint angles → compute end-effector pose
Inverse kinematics (IK)Given desired pose → compute required joint angles
Jacobian matrixMaps joint velocities to end-effector velocities: ẋ = J·θ̇
DH parametersDenavit-Hartenberg convention: a, α, d, θ per joint
WorkspaceReachable volume; regular vs dexterous workspace
SingularityConfiguration where manipulability = 0; Jacobian rank drops

ROS 2 Key Concepts

ConceptDescription
NodeBasic process unit; subscribes/publishes/provides services
TopicAsync publish-subscribe communication channel
ServiceSynchronous request-response between nodes
ActionLong-running task with feedback + preemption
TF2 (transform library)Track coordinate frame transformations over time
URDFXML format describing robot kinematics and visuals
ros2 launchLaunch multiple nodes from a single file
colcon buildBuild tool for ROS 2 packages