As companies rush to develop delivery robots, disinfection systems, and autonomous vehicles, their engineering teams must maximize productivity amid pandemic and business constraints. New software and hardware platforms put artificial intelligence at the edge to facilitate robotics innovation and speed to market. For instance, ADLINK Technology Inc. said its Edge AI portfolio automates edge computing processes to free developers from the need for advanced knowledge of data science and machine learning models.

In June, ADLINK released the ROScube-I, a new controller using Intel processors that it said supports a wide variety of sensors and actuators to enable faster, easier, and scalable development of open robotics applications. The Taiwan-headquartered company also launched the ROScube-X, which uses the NVIDIA Jetson AGX module for robotics applications requiring artificial intelligence and minimal power consumption. Both controllers are built on ROS 2, the latest version of Open Robotics’ Robot Operating System.

“The difference between autonomous machines such as an iRobot Roomba and an autonomous passenger vehicle is a matter of scale,” said Joe Speed, field chief technology officer at ADLINK. “It is a difference in the scale of the sensor suite, the in-vehicle compute, complexity, and risk.”

“Many passenger-carrying programs are stalled because of the coronavirus, but AMRs [autonomous mobile robots], AGVs [automated guided vehicles], cleaning robots, and autonomous cargo-delivery efforts are hugely accelerated,” he told The Robot Report. “There’s a 20x increase in robot orders in some cases, with companies accelerating two-year programs to delivery within six months.”

Defining edge AI and sharing more machine vision R&D

“We don’t use the mangled definition of ‘edge computing’ that mobile network operators use,” Speed said. “We’re not just talking about the edge of mobile networks; we’re moving compute to where sensors are and data is generated — not the office or the data center. These controllers are for computer vision and robots in industries like aerospace and vehicle manufacturing.”

“ADLINK’s technology and software partners enable things like optical inspection, hand-eye coordination for robots, and rugged industrial applications,” he said. “Cameras with edge AI compute can be built right into the end of the robot arm.”

“There are people doing a lot with computer vision with hardware-accelerated machine learning for object detection and SLAM [simultaneous localization and mapping], but not enough of that work gets open-sourced and contributed upstream to ROS,” Speed noted. “With ROS, we have a solid framework to accept such work, and some university work is published. But not enough machine learning work is reposted back to is contributed back to ROS/ROS 2 Github repos for perception, navigation, etc.”

“Intel has done well in ROS with its CPUs and OpenVINO, enabling the MoveIt path planner, and for perception, adding Movidius support for the RealSense cameras in ROS. With its Keem Bay modules, Intel claims it has achieved Xavier-level performance on a fifth of the power budget,” he said. “NVIDIA historically hasn’t done a lot of direct contributions to ROS, but I think that’s going to change with Jetson GPUs [graphics processing units] and contributions to ROS/ROS 2.”

ADLINK touts ties to the global robotics community

“Unlike random box builders, ADLINK has software labs around the world,” said Speed. “We’re big contributors to ROS 2, along with Amazon Web Services [AWS], Bosch, iRobot, and Open Robotics, as well as contributing to the Eclipse Foundation Eclipse IoT and OpenADx and the Autoware Foundation‘s open-source autonomous driving for ROS/ROS 2.”

“ADLINK has 1,800 people and its own factories,” he added. “We do a lot of manufacturing with Intel and NVIDIA, as well as NXP and Qualcomm.”

Speed said he expects big hardware and middleware players to increasingly support ROS. “Arm recently adopted the Eclipse Cyclone DDS [Data Distribution Service] ROS middleware and plans to work on profiling, optimization, and building in its TrustZone hardware security into its DDS Security,” he said. “Eclipse Cyclone DDS is popular Tier 1 middleware built into ROS 2 and Autoware. It is easy to use, quite small, very fast, and extremely efficient and reliable.”

“It can move data using half as many CPU cycles — this has implications for the battery life of mobile robots,” said Speed. “Its open-source contributions have come from many developers, including ones at ADLINK, Rover Robotics, Ekumen, Canonical, and AWS.”

ADLINK says ROS 2 designed for commercial developers

ROS was widely viewed as not robust enough for commercial applications, but the Foxy Fitzroy build presents new opportunities, Speed said.

“A lot of people may have tried ROS in the past and didn’t like it,” he acknowledged. “ROS ‘classic’ was very much perceived as a tool for academics, so companies would harden it and add to it as they built their own products on it. They didn’t advertise it, but many things like Cruise Automation were originally built on ROS.”

“It’s not obvious to outsiders, but ROS 2 has a heavy investment from big technology corporations,” said Speed. “There are dozens of contributing engineers at AWS, Samsung, LG, and others contributing daily. Most are not roboticists but are systems architects, software engineers, and testing and QA [quality assurance] people with serious skills. This is why we’re seeing such rapid and constant improvements in the ease, performance, and reliability of ROS 2.”

“We knew what is needed for ROS 2. It has to be fault tolerant, real-time, and functionally safe,” he said. “The ROS community re-architected ROS to be built on DDS, which is used by industry, the military, and aerospace. ADLINK-contributed DDS is in every Fujitsu Internet backbone fiber switch, the ship-defense systems of 18 navies, autonomous vehicles, and industrial and farming robots.”

“By building robots on a solid foundation with open-source middleware, we can all get to autonomy faster,” Speed claimed. “Engineers can get to bigger, heavier things that move faster, with more valuable cargo including people.”

“If you were on the fence about trying open robotics or ROS 2, now is the time with Foxy,” he said. “This release has the contributions of 115 engineers. If you look at the development and testing for this open-source release, it would have been an $8 million commercial software effort.”

“I challenge you to find more than 10 robotics startups that are not using ROS or ROS 2. Most are using it,” said Speed. “ROS is coming to robot arms, too. Trajekt Sports has developed a robot that can accurately replicate any baseball pitch thrown in a MLB game — everything from curveballs to 105+ mph fastballs.”

Pandemic puts demands on robotics development

“While autonomous passenger vehicle programs have slowed a bit or are stalled, delivery robot programs have sped up,” said Speed. “For example, Damien Declercq at Spring Mobility told me how at Neolix has seen a 20x increase in delivery robot orders in 90 days because of COVID-19. The trends remain the same, but the pandemic has changed the timelines, elongating some and compressing others.”

“I have friends at Box Robotics in Philadelphia who have 36 years of experience in warehouse robots,” he said. “They say COVID-19 has not only accelerated the shift from retail stores to e-commerce, but it has also put a real strain on warehouses and supply chains.”

“We’re working with Ouster and SICK on using ROS 2 perception to make a warehouse vehicle drive twice as fast safely. This would save 30% of the cost and get more throughput,” Speed said. “Most things move at 2 m/sec. but really at 1 to 1.5 m/sec., but they should really be able to move the speed of a human or forklift at 5 m/sec.”

“Box Robotics, ADLINK, and AWS RoboMaker are working with Ouster and SICK on using ROS 2 perception to make autonomous warehouse vehicles drive twice as fast safely,” said Speed. “This would save 30% of the cost and get more throughput. Most autonomous things in warehouses move at 1 to 1.5 m/sec., compared with a human-driven forklift at 5 m/sec.”

ADLINK’s customers include a major aerospace manufacturer that requires any robots that it buys to support ROS, Speed said. “ROS Industrial is starting to be built into the procurement requirements of big manufacturers,” he said.

Fleet management challenges ahead

“There is a multiyear program to automate Changi General Hospital in Singapore,” he said. “It’s a heterogenous environment in which robots must integrate with hospital systems and multiple robots. A robot can summon an elevator, roll into it, and make way for another robot. These robots are already sharing space with humans, doing everything from resupply to bedside care, all working together.”

ADLINK is also working with providers of autonomous delivery, AMRs, and self-driving vehicles, but Speed declined to name any partners.

“With the Autoware Foundation developing support for autonomous cargo delivery, a robot could pick up parts in one building of a car factory and navigate onto a street and into another building to deliver parts to the manufacturing line at the moment they’re needed,” he said. “This combines indoor and outdoor navigation and operating in areas with other robots and humans.”

“ADLINK got me to join in large part because of its support for assistive technologies,” recalled Speed. “My first effort upon joining ADLINK was delivering #AccessibleOlli, an open-source, self-driving bus for elderly and disabled. This was a #tech4good community effort with contributions from MIT, Princeton, AARP, IBM, the Mayo Clinic, and many more. My team of P-TECH kids chose Stevie Wonder as their design persona. So it was pretty great when our friend Mike May brought Stevie Wonder to experience what my kids had designed for him.”