Keep learning
Resources
A small toolkit for reading, observing and asking better engineering questions.
Linux observation checklist
These commands inspect a Linux system without changing its configuration. Output and available tools vary by distribution. Review logs and process information before sharing them.
uname -r # Kernel release
uptime # Uptime and load averages
free -h # Memory and swap
df -h # Filesystem use
systemctl --failed # Failed systemd unitsA single snapshot is not a performance test. Write down the workload, time and software versions before drawing conclusions.
Download plain-text checklist ↓Robotics glossary
- LiDAR
- A sensor that uses light to measure distance; multiple measurements can form a scan.
- IMU
- An inertial measurement unit, commonly combining accelerometers and gyroscopes.
- Point cloud
- A set of spatial measurements represented as points.
- SLAM
- Simultaneous localization and mapping: estimating pose while building a map.
- Odometry
- An estimate of movement relative to an earlier position.
- Sensor fusion
- Combining observations with a model and uncertainty assumptions.
- Coordinate frame
- A reference origin and set of axes used to describe a measurement.
- Covariance
- A representation of uncertainty and correlations between quantities.
- Inference
- Running a trained model on a new input.
- Quantization
- Representing model values with reduced precision.
- Latency
- Elapsed time between a defined input event and its corresponding result.
- Throughput
- The amount of work completed per unit of time.
- Jitter
- Variation in timing; always state which timing interval is being measured.
- Deadline
- The latest acceptable completion time for a task.
- Torque
- A turning moment, measured in newton-metres.