olive™ AI CAMERA

camera

olive™ Camera is a software-defined camera for robots designed to be re-programmable and easy to integrate allowing you to quickly create custom robots.

  • Modular and software-defined
  • Plug & play
  • Native ROS & ROS 2 interface
  • Built-in AI-accelerator for edge AI
  • Real-time Kernel
  • Ethernet over USB connection
Buy olive™ Camera Datasheet

DESIGNED AND
MANUFACTURED
IN GERMANY 🇩🇪

Robot hardware that grows with you

Learn, build and grow in robotics with Olive™ hardware.
Create robots intuitively while aligning with industry standards.

RE-PROGRAMMABLE

REAL-TIME

INTEROPERABLE

AI ACCELERATOR

ROS & ROS 2-NATIVE

Ehternet USB-C

Main Features

WebGUI: The Olive embedded web-based GUI is a user-friendly interface for configuring, controlling, and testing out our modular robotic building blocks. It provides a wide range of capabilities for setting up and controlling our hardware components, including setting parameters, sending commands, and visualizing sensor data. It also offers an Integrated Development Environment (IDE) that is based on the JupyterLab platform. It allows users to manage embedded files on the device, create projects, write Python or C++ based embedded applications, and run them on the modules.
Tensor Processing Unit (TPU): A Tensor Processing Unit (TPU) is a custom-designed application-specific integrated circuit (ASIC) developed by Google to accelerate machine learning workloads. Unlike traditional CPUs and GPUs that manage a broad range of tasks, TPUs are optimized for the high-speed matrix multiplications and computations central to neural network operations. Their architecture offers substantial advantages in terms of performance per watt and computational throughput, making them particularly effective for training and inferring large-scale machine learning models, especially within Google's machine learning framework, TensorFlow.
Integrated IMU: The Olive camera comes with an integrated IMU, crucial for visual SLAM algorithms. While cameras provide detailed visual data, they can experience drift or occlusions. The IMU offers high-frequency orientation and acceleration information. By merging both data sources, accuracy and resilience against visual ambiguities improve, leading to a more reliable estimation of camera movement and scene mapping.
ROS/ROS 2 Native: Embracing the advancements of modern sensor technology, our Camera offers native support for ROS2 (Robot Operating System 2). This essentially means that the Camera is driverless, eliminating the need to install and run a separate ROS2 node on your host computer. By sidestepping this additional layer of software, not only is the system latency minimized, but the throughput of data is also optimized to its maximum potential. This direct integration ensures that data flows seamlessly and promptly from the sensor to the system. Furthermore, the native ROS2 capability makes the Camera truly plug-and-play, streamlining integration processes and enhancing user experience, especially in dynamic environments where rapid deployment and configuration are crucial.

A universal robot camera

Modular. Software-defined. ROS and ROS 2-native

olive™ Camera is a revolutionary modular robot camera that
allows you to build any robot fast and easily.

  • Real-time kernel: A real-time kernel offers deterministic response times, making it essential for consistent data acquisition. Its optimized nature minimizes system latencies, ensuring timely processing in critical applications. This predictability is key to avoiding unexpected delays or interruptions.
  • ROS & ROS 2 Native Perception: This essentially means that the Camera is driverless, eliminating the need to install and run a separate ROS2 node on your host computer.

  • TPU: a specialized chip designed for accelerating neural network computations. This integration allows us to run advanced neural network models directly within the camera, enhancing its capabilities for real-time image analysis and processing.
  • Metal Case: Ensures swift heat dissipation, maintaining the device's optimal operating temperature. This robust design not only enhances performance but also extends the device's longevity by effectively combating thermal stress.

Camera tech specs

Powered by ROS

Olive™ Camera has built-in ROS capabilities. The Robot Operating System (ROS) is an open-source framework that enables developers to easily create, manage, and integrate robot hardware components. It is widely used in the robotics industry, and is considered a crucial tool for anyone working in the field of robotics. ROS provides a common set of tools and libraries that can be used to develop and control robot hardware, making it possible for hardware components from different manufacturers to work together seamlessly. This is particularly important for the development of modular robots, which are composed of multiple interchangeable parts. At Olive Robotics, we use ROS and ROS 2 to develop and control our robot hardware components, enabling us to create advanced, modular robots that are easy to program and customize.

Which companies are using ROS?

Plug and play,
ROS-native hardware

olive™ components is designed to be easily integrated into any robotics system, thanks to its plug and play design. This means that it can be easily connected to a computer or control system using a simple USB-C interface, without the need for complex wiring or configuration.

In addition, olive™ component is natively compatible with ROS, Gazebo and all of its ecosystem of tools. This allows users to easily incorporate Olive™ components into ROS-based projects in an accelerated manner.

Applications

Empowering everyone to sense motions, reliable, plug and play

Robotic Agriculcture
Robotic Warehouse
Robotic Inspection
Autonomous Cars
Indoor Service Robots
Traffic Monitoring
Last Mile Delivery Robots
3D SLAM Systems
Teleoperation
Surveillance Systems
Industrial Quality Control
Robotic Manipulation

TECH SPECS

Form Factor

Dimensions (W x H x D)

Weight

40mm x 40mm x 30mm

89 grams

Processor Unit

Application Processor

Real-Time Processor

TPU AI Accelerator

Dual Cortex-A7 up to 800 MHz

Cortex-M4 MPU up to 200 MHz

4 Trillion Operations Per Second

memory

On Chip (SoC)

On Chip EEPROM (SoC)

On SOM

512 MB RAM

512 Bytes x 8

64 GByte SD Flash

Sensors

Image Sensor

Max Resolution

Lens

1/2.7" OmniVision OV2710

1920(H)x1080(V) pixels

2.1mm (L210) / 1.8mm (L180)

Frame Rate

640x480 VGA @120fps

1280x720 HD @60fps

1920x1080 FHD @30fps

IMU Sensor

6-Axis Automotive-Proven IMU

- Accelerometer: 0.06 mg/LSB

- Gyroscope: 0.004 dps/LSB

IMU Range & Sensitivity

Output Data Rate (ODR): 1000 Hz

Accelerometer Zero-g ±20 mg

- Range ±16 g

- Sensitivity 2048 LSB/g

Gyroscope Zero-rate ±0.5 dps

- Range ±2000 dps

- Sensitivity 16.4 LSB/dps

Connectivity

High-Speed Connectivity

1x Virtual Ethernet USB Type C

@ 60 Megabytes Per Second (MBps)

Other I/O

1x User Switch

3x User LEDs

Software

Yocto BSP available for batch purchases

Linux Kernel

Operating System

Communication middleware

Robotic Operating Systems

Linux 5.10

Debian 11

Apache Cyclone DDS

ROS Noetic Ninjemys, ROS 2 Humble Hawksbill

Power and Thermal

Power Consumption

Voltage

Max Current

Temperature Range

USB Type C PD (15 W max)

PD 5.0v

3000mA

Commercial 0°C to 85°C / Industrial -40°C to +85°C

Form Factor

Dimensions (W x H x D)

40mm x 40mm x 30mm

Weight

89 grams


Processor Unit

Application Processor

Dual Cortex-A7 up to 800 MHz

Real-Time Processor

Cortex-M4 MPU up to 200 MHz

TPU AI Accelerator

4 Trillion Operations Per Second


Memory

On Chip (SoC)

512 MB RAM

On Chip EEPROM (SoC)

512 Bytes x 8

On SOM

64 GByte SD Flash


Sensors

Image Sensor

1/2.7" OmniVision OV2710

Max Resolution

1920(H)x1080(V) pixels

Lens

2.1mm (L210) / 1.8mm (L180)

Frame Rate

640x480 VGA @120fps

1280x720 HD @60fps

1920x1080 FHD @30fps

IMU Sensor

6-Axis Automotive-Proven IMU

- Accelerometer: 0.06 mg/LSB

- Gyroscope: 0.004 dps/LSB

IMU Range & Sensitivity

Output Data Rate (ODR): 1000 Hz

- Accelerometer Zero-g ±20 mg

- Gyroscope Zero-rate ±0.5 dps


Connectivity

High-Speed Connectivity

1x Virtual Ethernet USB Type C

@ 60 Megabytes Per Second (MBps)

Other I/O

1x User Switch

3x User LEDs


Software

Yocto BSP available for batch purchases

Linux Kernel

Linux 5.10

Operating System

Debian 11

Communication middleware

Apache Cyclone DDS

Robotic Operating Systems

ROS Noetic Ninjemys, ROS 2 Humble Hawksbill


Power and Thermal

Power Consumption

USB Type C PD (15 W max)

Voltage

PD 5.0v

Max Current

3000mA

Temperature Range

Commercial 0°C to 85°C / Industrial -40°C to +85°C


Partners

Projects and customers all around the world

Do you have any questions?

Get in touch with our team.

Let's talk