Refresh

This website linuxgizmos.com/ubuntu-ros-robotics-sbc-has-rpi-style-gpio-and-a-choice-of-atom-or-rockchip-px30/ is currently offline. Cloudflare's Always Online™ shows a snapshot of this web page from the Internet Archive's Wayback Machine. To check for the live version, click Refresh.


All News | Boards | Chips | Devices | Software | LinuxDevices.com Archive | About | Contact | Subscribe
Follow LinuxGizmos:
Twitter Facebook Pinterest RSS feed
*   get email updates   *

Ubuntu/ROS robotics SBC has RPi-style GPIO and a choice of Atom or Rockchip PX30

Jun 9, 2020 — by Eric Brown 1,487 views

Adlink announced a “ROScube Pico Development Kit” SBC with 40-pin GPIO and a Myriad X VPU that runs Ubuntu and ROS 2 on an Apollo Lake or Rockchip PX30. There is also a Lidar-equipped, 6th Gen Skylake based NeuronBot robot.

Earlier today, we looked at Adlink’s ROScube-I robot controller and ROScube-I Starter Kit SBC based on Intel Coffee Lake processors. Part two of that announcement concerned a new ROScube Pico Development Kit and NeuronBot robot that we cover here.



ROScube Pico Development Kit and NeuronBot
(click images to enlarge)

Like the ROScube-I products, the ROScube Pico and NeuronBot run Ubuntu 18.04 LTS along with Adlink’s Neuron SDK, which based on the latest Foxy Fitzroy version of ROS (Robot Operating System) 2 middleware. The products also support Intel’s OpenVINO toolkit for AI.

 
ROScube Pico Development Kit

— ADVERTISEMENT —


The “preliminary” ROScube Pico Development Kit, which is referred to as the ROScube-I Pico Development Kit in the announcement, will be available in two different processor SKUs, each deployed via Adlink’s LEC-branded SMARC modules. The Apollo Lake based NPS-1 and the Rockchip PX30-based NPS-4 models appear to be based loosely on Adlink’s previously announced Vizi-AI dev kit, based on Adlink’s LEC-AL module. Adlink also launched a very similar I-Pi SMARC Dev Kit based on the LEC-PX30.



Adlink’s Vizi-AI (left) and I-Pi SMARC Dev Kit
(click images to enlarge)

The preliminary ROScube Pico product page refers to the processor as simply an Intel Atom. It is likely an Apollo Lake model since the later Gemini Lake generation that follows has yet to release a model using the Atom name. The LEC-AL module used on the Vizi-AI is the quad-core Atom x5-E3940. The ROScube Pico NPS-1 is also equipped with an Intel Myriad X VPU for AI processing.

The LEC-PX30, I-Pi SMARC Dev Kit, and ROScube Pico NPS-4 all use the quad-core, Cortex-A35 Rockchip PX30. The PX30 SoC provides a Mali-G31 MP2 GPU and a video processor that supports dual independent HD displays. The NPS-4 lacks the Myriad X, which is probably implemented via the SBC’s an M.2 E-key slot, but could also be the external device on the bottom shelf of the acrylic case in the image below.



ROScube Pico side view
(click image to enlarge)

The 118 x 81.8mm ROScube Pico is less feature rich than the Vizi-AI and I-Pi boards. The NPS-1 version of the SBC ships with 8GB LPDDR4 and 32GB eMMC while the NPS-4 has 2GB DDR3L with no flash. Both models ship with a microSD slot with optional preconfigured card.

The ROScube Pico is equipped with a GbE port, 2x USB 3.0, and 2x USB 2.0 ports. Media features include an HDMI port, MIPI-DSI and -CSI connectors, and a headphone jack. There is also a Raspberry Pi compatible 40-pin GPIO connector.

The ROScube Pico has a 12V jack, a 24W AC/DC adapter, and power and reset buttons. Unlike the Vizi-AI and I-Pi, the SBC provides ruggedization features including 0 to 60°C support, 1Grms vibration resistance per IEC60068-2-64, and 50G shock resistance per IEC-60068-2-27.

 
NeuronBot

The NeuronBot Series is an autonomous, dynamic motion robot development platform designed for research, training and education. The robot runs Ubuntu and the Neuron SDK on a dual-core, 2.3GHz Celeron G3900TE from Intel’s 6th Gen Skylake family.

The dual wheel-equipped, 260 x 270 x 260mm robot weighs 8.3 kg and can carry payloads up to 3 kg. An Arduino Mega 2560 board controls a dual-channel 7N14P encoder and DC Carbon-brush motor/actuator for motor control.

There is a connector for an optional Intel RealSense D435 depth camera and a laser distance Lidar camera. The 2D, 360° RPlidar A1 Lidar has a 12-meter range, 8000Hz sample frequency, and a 1-10Hz scan rate range. It is connected via a USB/UART interface. The robot has 0.6m/s rotational and translational velocity and a 0 +/- 1° climbing threshold.

The NeuronBot ships with 4GB DDR RAM and a 64GB SSD. Major ports include 2x GbE, 5x USB 3.0, 4x USB 3.0, and HDMI and RS-232/422/485 ports. Internally, there are 2x USB 3.0, 3x RS-232, 10x GPIO, and an I2C interface.

Other features include a GY85 9-axis IMU, a 12V DC input with ATX connector, and a power button. There are 4x LEDs and a status display for the optional 22.2V/3300mAh LiPo battery. A “flat panel” is also optional. Ruggedization features include a 0 to 50°C range, IEC 60068-2-64 compliant vibration resistance, ISTA-1A rated drop resistance, 10%~95%, non-condensing humidity tolerance, and CE and FCC Class B EMC certifications.

 
Further information

No pricing or availability information was provided for the “preliminary” ROScube Pico Development Kit or the “new” and “affordable” NeuronBot. More information may be found in Adlink’s announcement and the ROScube Pico and NeuronBot product pages.

 

(advertise here)


Print Friendly, PDF & Email
PLEASE COMMENT BELOW

Please comment here...