-
Features
-
Specification
-
Files
Object Detection & Counting
Object detection and counting are essential for various applications, including inventory control, production line monitoring, and shipment verification. Traditional methods, such as manual inspection and mechanical counting, are often slow, labor-intensive, and susceptible to errors. Machine vision systems offer a compelling alternative, providing rapid and reliable object detection and counting. Their speed and accuracy significantly outperform manual and mechanical methods, leading to improved efficiency and reduced errors in these critical processes.
Locating Guiding & Positioning
Many automated processes, including assembly, pick-and-place operations, and inspection, rely on machine vision's ability to precisely locate parts, regardless of whether they are on a conveyor or in a bin. Machine vision systems guide robots in these tasks, enabling accurate part placement and even micron-level precision in assembly. Furthermore, if a part is misaligned, the system can measure the discrepancy between the desired and actual position and orientation, then communicate these adjustments to a robot or PLC for corrective action.
Defect Detection
Defects can arise at any point in manufacturing, from raw materials to final inspection. While traditionally performed manually, inspections are now significantly improved by machine vision. Operating at production line speeds, these systems offer consistent, tireless performance, detecting even minute and unexpected defects while also capturing data for continuous improvement, advantages unmatched by human inspectors.
OCR/OCV & Barcode Reading
Supports all character types and prints, including embossed, engraved, laser-marked, and inkjet-printed parts. Detection remains reliable regardless of target variations or environmental conditions.
Active Stereo Depth
Active Stereo Depth uses a dot projector to enhance disparity matching, improving accuracy on low-detail surfaces. While the matching process is similar to Passive Stereo, the projected dots make a critical difference.
Night Vision
Infra Red LED illumination enables clear vision in low-light or no-light conditions. It supports AI/CV processes by providing IR-lit frames, ideal for mono camera streams, as color cameras don't detect IR light.
DIN Rail Mount
DIN Rail Mount option provides a secure and space-saving solution for integrating the PLC controller into industrial control panels or enclosures. This standardized mounting method allows for easy installation and organization alongside other DIN rail compatible equipment, streamlining your setup and reducing clutter.
Rugged Body
Featuring an IP65 rating for protection against dust and water ingress, a robust aluminum body to withstand impacts, and scratch-resistant Gorilla Glass for a crystal-clear view, the camera is ready for anything. The inclusion of an industrial-grade M12 Ethernet connector ensures a secure and stable network connection even in harsh conditions, making the camera ideal for demanding industrial applications.
Linux Based OS
Experience the power and flexibility of a Linux-based operating system with our advanced PLC controller. This open-source platform provides unparalleled customization options, allowing you to tailor the controller's functionality to your exact needs. Benefit from a vast community of developers and readily available resources, ensuring seamless integration with other Linux-compatible systems and software. The robust and stable nature of Linux translates to reliable performance and minimal downtime, making our Linux-based PLC ideal for complex automation tasks and demanding industrial environments.
Works Offline
Our Machine Vision System offers powerful image processing and analysis capabilities, even without a constant network connection. Designed for reliable performance in any environment, this system operates seamlessly offline, eliminating concerns about network interruptions or latency. This standalone functionality ensures consistent and uninterrupted inspection, measurement, and identification, making it ideal for remote locations or applications where real-time connectivity is unreliable or unavailable. Process images, analyze data, and make critical decisions directly on the device, maximizing efficiency and minimizing downtime.
Controller
| Digital Inputs | 4 |
| Digital Outputs | 4 |
| Analog Inputs | 1 |
| Analog Outputs | 1 |
| CPU | 4× Arm© Cortex©-A53, max 1.8 GHz |
| RAM | 1 GB |
| Internal Storage | 8 GB |
| LTE Module | 1 |
| RS485 Interface | 2 |
| LTE Modem | Quectel EG912Y-EU |
| Ethernet | 100 Mbit |
| USB 2.0 | 2× |
| Power | 24 V DC |
| 1-Wire Bus | Yes (galvanically isolated) |
| Mounting | DIN rail (35 mm) |
| Ingress Protection | IP20 |
| Dimensions | 70 mm × 90 mm × 60 mm |
| Weigth | 198 g |
Digital inputs (DI) are designed for reading states of two-state devices such as various digital sensors, switches and buttons, motion sensors etc. All digital inputs feature counter functionality by default.
Digital outputs (DO) are applicable for switching two-state devices using direct current (DC) such as locks, external relays, contactors, etc. DO’s can be also used for analog control thanks to the pulse-width modulation (PWM) function.
Analog input (AI) is designed to measure voltage or current in order to read values from various analog sensors (pressure meters, tension gauges, etc.).
Analog output (AO) is used to control and regulate devices such as three-way valves, electric motors, inverter drives, heat exchangers, etc. by direct current or voltage.
Computing Module
Controller CPU is based on the NXP i.MX 8M Mini platform featuring a quad-core ARM Cortex-A53 CPU with 1.8 GHz maximum frequency and 1 GB LPDDR4 RAM. This makes the controller the most powerful compact PLC available on the market. Internal 8 GB eMMC memory offers higher reliability compared to systems utilizing microSD memory cards.
GSM Modem
The controller features an LTE modem and a microSIM card slot to connect to the LTE wireless network.
| Chip model | Quectel EG912Y-EU |
| LTE category | Category 1 |
| Download speed | up to 10 Mb/s |
| Upload speed | up to 5 Mb/s |
| Frequency Bands | B1/B2/B3/B5/B7/B8/B20/ B28/B38/B40/B41 |
The modem allows the controller to serve as a primary or secondary internet access point for remote management and data transmission through the internet, or for sending and receiving SMS messages. In the role of a primary connection the controller functions as a web access point in locations without a landline internet connection. Secondary connection then comes into play as a backup in case of a primary Ethernet connection failure in technologies reliant on a connection between the technology and a supervision center. In the case of Ethernet failure, the controller automatically switches to the LTE mode, bridging the connection.
Note: The product requires a standard microSIM (0.76mm thickness) card. NanoSIM cards (0.67mm thickness) are not supported.
Communication Interfaces
Controller feature RS485 serial lines by default. This type of interface usually utilizes the Modbus RTU protocol and can be used for communication with various devices such as energy meters, touchscreen HMI displays, controllers from other vendors, various sensors, etc. A single bus can support dozens of devices while its length can reach up to several hundred meters.
The controller also feature a single RS232 serial line. In the field of automation, this interface is usually used for communication with various converters, closed-in technologies, or display devices such as HMI panels.
A single 1-Wire bus is included as a standard feature. This interface uses 2-3 conductors for collecting data from various 1-Wire sensors such as thermometers, humidity meters, light intensity sensors, etc. A single 1-Wire bus supports up to 15 devices, the overall length of the bus can reach up to 200 meters.
Controller also feature a 100 Mbit Ethernet port for network communication.
Camera
| Processing Power | 4 TOPS (1.4 TOPS for AI) |
| Active Stereo | IR dot projector |
| Night vision | IR illumination LED enables running your AI and CV algorithms in low-light or no-light environments |
| Run any AI model | |
| Encoding | H.264, H.265, MJPEG - 4K/30FPS, 1080P/60FPS |
| Base line | 75 mm |
| Ideal depth range | 70cm - 12m |
| Power consumption | Up to 7.5W |
| Ingress Protection | IP65 |
| Mount | 1/4" tripod mount and 75mm M4 Vesa mount |
| Dimensions | 111x40x31.3mm |
| Weight | 184g |
| Manufacturer | Luxonis |
Available in three variants
The OAK-D Pro Poe is available in three RGB camera variants: standard versions IMX378 Fixed-Focus and IMX378 Auto-Focus and special version OV9782 Fixed-Focus. The Fixed-Focus variant has the lens fixed in place mechanically, whereas the Auto-Focus has an electromagnet that controls the lens position dynamically. All variants can clearly see objects 50 cm and beyond.
- Auto-Focus cameras are recommended for close objects between 8 cm to 50 cm of the device (see video). They can also be manually set to a specific focus position utilizing the DepthAI API.
- Fixed-Focus cameras are recommended for heavy-vibration applications, such as drones, lawnmowers, or vehicles (see video).
- OV9782 Fixed-focus camera feature a special 120FPS global shutter sensor OV9782
Sensors Specifications
| Color camera | Stereo Pair | |||
|---|---|---|---|---|
| Variant | Auto-Focus | Fixed-Focus | OV9782 | All |
| Sensor | IMX378 (PY004) | IMX378 (PY052) | OV9782 (PY074) | OV9282 (PY044) |
| FOV (D/H/V) | 78° / 66° / 54° | 82° / 69° / 55° | 89.5° / 80° / 55° | 89.5° / 80° / 55° |
| Resolution | 12MP (4056x3040) | 12MP (4056x3040) | 1MP (1280x800) | 1MP (1280x800) |
| Focus | AF: 8cm - ∞ | FF: 50cm - ∞ | FF: 19.6cm - ∞ | FF: 19.6cm - ∞ |
| Max Framerate | 60 FPS | 60 FPS | 120 FPS | 120 FPS |
| Spectral range | Visible + 940nm | Visible + 940nm | Visible + 940nm | Visible + 940nm |
| Shutter Type | Rolling | Rolling | Global | Global |
| Sensor size | 1/2.3” | 1/2.3” | 1/4” | 1/4” |
| F-number | 1.8 ± 5% | 1.8 ± 5% | 2.0 ± 5% | 2.0 ± 5% |
| Effective Focal Length | 4.81mm | 4.81mm | 2.35mm | 2.35mm |
| Pixel size | 1.55µm | 1.55µm | 3.0µm | 3.0µm |
Stereo depth perception
This OAK camera has a baseline of 7.5cm - the distance between the left and the right stereo camera. Minimal and maximal depth perception (MinZ and Max) depends on camera FOV, resolution, and baseline - more information here.
- Ideal depth range: 70cm - 12m
- MinZ: ~20cm (400P and extended), ~35cm (400P OR 800P, extended), ~70cm (800P)
Depth accuracy (See 800P, 75mm baseline distance OAKs for details):
- below 4m: below 2% absolute depth error
- 4m - 7m: below 4% absolute depth error
- 7m - 10m: below 6% absolute depth error
IR perception
This product is classified as a Class 1 Laser Product under the EN/IEC 60825-1, Edition 3 (2014) internationally. Test protocol from laser class classification can be downloaded from here. More information about IR perception and laser safety can by found here.
Active stereo:
- Uses Ams Belago 1.1 Dot-Pattern Infrared Illuminator
- IR dot projector improves depth perception, especially for low-visual-interest surfaces by projecting thousands of IR dots to the scene.
Night vision:
- IR illumination LED enables running your AI and CV algorithms in low-light or no-light environments
Rugged industrial connectors
M12 PoE connector is used for power and 1Gbps data transfer. Additionally, M8 connector for IO connectivity; USB2 (D+, D-), camera IOs (FSIN, STROBE), power, and one auxiliary GPIO. Cap for M8 is included when connector is not used to ensure water resistance. More information can be found here.
Rotate the locking ring (not the connector body) clockwise (CW) with <0.8 Nm torque to securely connect the cable.
Integrated IMU
This OAK camera has an integrated BNO085, a 9-axis IMU (Inertial Measurement Unit). See IMU node for the API details on how to use it.
Power-over-Ethernet (PoE)
The base board offers full 802.3af, Class 3 PoE compliance with 1000BASE-T speeds (1Gbps). PoE injector/switch is required to power the device.
Power consumption
Most of the power is consumed by the RVC2, so the power consumption mostly depends on the workload of the VPU:
- Base consumption + camera streaming: 2.5W - 3W
- PoE circuitry: 0.5W
- AI subsystem consumption: Up to 1W
- Stereo depth pipeline subsystem: Up to 0.5W
- Video Encoder subsystem: Up to 0.5W
Illumination LED and laser dot projector
Power is configurable (more power - more brightness):
- Dot projector (active stereo): Up to 1W
- Illumination LED (night vision): Up to 1W
So the total power consumption can be up to ~7.5W if you are using all the features at 100% at the same time. To reduce the power consumption, you can reduce FPS of the whole pipeline - that way, subsystems won’t be utilized at 100% and will consume less power.
Water resistance
This camera is shipped with IP65 rated enclosure. More details can be found here.
Temperature
The ambient operating temperature of RVC2 based devices is between -20°C and 50°C when fully utilizing the VPU. To find out more, see our operating temperature range documentation.
Vibration resistance
Our cameras are tested for vibration tolerance according to the EN 60068-2-6:2008 standard. More information about the testing you can find here.
Software
This open-source machine vision kit provides a robust and flexible platform for industrial automation and intelligent control. Built on the stability and versatility of Linux, it offers a modern approach to automation, seamlessly integrating traditional PLC functionality with the power of neural networks.
-
Linux Based: Leverages the reliability and extensive software ecosystem of Linux, providing a stable and customizable foundation.
- REST API: Enables easy integration with various systems and platforms for monitoring, control, and data exchange. Ideal for web-based applications and general-purpose communication.
- WebSocket API: Facilitates real-time, bidirectional communication, essential for low-latency applications and responsive control systems.
- Input/Output Control: Direct access and control of the controller's inputs and outputs via both REST and WebSocket APIs, allowing for flexible integration with sensors, actuators, and other hardware.
- Neural Network Deployment: Deploy and execute custom neural networks directly on the controller, enabling advanced data processing, pattern recognition, and intelligent decision-making within your automation applications. This feature empowers the creation of sophisticated AI-driven control systems.
- Open Source: The entire platform is open source, fostering community involvement, transparency, and customization.
Resources:
- OS Image: Download the Debian-based OS image from GitHub: https://github.com/Robopipe/OS
- API Source Code: Explore and contribute to the API source code on GitHub: https://github.com/Robopipe/API
- Documentation: Access comprehensive documentation, including tutorials, API references, and examples on GitBook: https://robopipe.gitbook.io/doc
Package contains
Camera
- OAK-D Pro PoE camera
- M12 plug to RJ45 cable, 2m
- M8 plug male to male cable, 2m
- Cleaning cloth
Controller
- Controller with pre-installed OS
- Removable screw terminals
- DIN rail holder (35 mm)
- 3× DIN rail holder mounting screw
- 1× grounding screw with a fan washer
- Antenna with magnetic mount
Note: PoE injector / switch is required to power the camera and it is not included in the packaging.
