FEATURE Robotics & Motion Control
Advancing robotics with AI-fi rst approach
I
ndustrial leaders of all stripes are eager to achieve the effi ciency, automation and enhanced security that Industry 4.0 promises to
deliver. However, for many, fundamental challenges still exist when trying to perform business operations and automate workfl ows across hybrid network environments and disparate systems. 5G and advances in artifi cial intelligence (AI) promise to change that, creating opportunities that will allow industrialists to explore more automated solutions in predictive maintenance, analytics to support against larger datasets, and reliable and safe human-robot collaboration.
The technical problems facing Industry 4.0
Expanding the range of applications for robotics in IIoT environments will require resolving long-standing technical roadblocks. The hype around 5G has driven high expectations for Industry 4.0. Businesses expect it to provide a plethora of solutions such as preventing production ineffi ciencies and running predictive maintenance. For this to be possible they will need solutions that are highly cognitive and extremely autonomous. At the same time, large machines that work with sensitive materials or in close proximity with people must be able to communicate in real time in order to mitigate any risk or danger that they may cause. While businesses rely on manual and cloud computing methods, it’s unlikely they will be able to appreciate the full benefi ts of Industry 4.0 any time soon. The base data computing requirements for machines is already high, and to ensure safety, determinism, security and AI enablement, will need to be built into embedded systems. AI combined with
automationmagazine.co.uk
new communication technologies will help integrate robotics into software- defi ned infrastructure. As businesses start to hand the computing work over to AI in intelligent edge devices, the rollout of robotics in industrial applications will become more mainstream. According to Deloitte, AI will be able to help plan driverless vehicle routes, save time and costs in supply chain management, increase reliability and analyse big data.
AI first For the robotics industry to thrive and address these problems head on, machines need to be built with an AI-fi rst approach. The rising interest in AI-fi rst solution modelling suggests that a rethinking of conventional industrial robotics design and development is needed. Rather than relying on an approach in which existing machine operations are essentially augmented by bolting on AI-driven components, AI-fi rst puts the intelligence at the forefront of the design process to perform at the core of a task. It should be noted that we are not implying that existing models are not valid nor eff ective, rather that they need to be augmented and updated to include AI beyond the notion of a “bolt-on”.
It’s expected that the number of robots deployed in the US by 2023 will exceed 400,000
Integrating AI and robots for industrial applications will generate nearly $66.5bn by 2027, writes
Michel Chabroux, Senior Director for Product Management at Wind River
The AI-fi rst approach opens the door to exciting applications in robotics. AI- enabled collaborative robots, autonomous vehicles, non-piloted drone operations, are just a few examples of an expanding array of innovative uses. Exoskeletons for example are an area of robotics that is gaining momentum from emerging technologies. They are designed to support heavy-duty tasks or ergonomically- challenging human operations. Since it works quite literally in collaboration with humans, an exoskeleton needs AI to assist its interactions with its human pilot. One of the most famous current examples of AI working with robotics is the Mars Perseverance rover. Over 130 million miles from Earth, having AI at the intelligent edge is critical to the success of unmanned space missions. Perseverance has to make decisions and actions almost entirely autonomously; the distance is much too far for humans to oversee all its decisions. For mission success, it must act on its own assessments, i.e. analysis of data, to determine its behaviour such as where to land, which routes to follow and what pictures to take.
A hot-ticket item The demand for robots and automated systems is rapidly growing, and this will drive innovation further. And, as AI becomes increasingly sophisticated and the intelligent edge is fortifi ed with new technologies, the opportunities and applications for industrial robots in manufacturing, transportation and aerospace will continue to multiply.
CONTACT:
Wind River
www.windriver.com
Automation | October 2021
19
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50
