Article

AI-powered robots coming to a factory near you

generic graphic of a headshot
man and woman watching robotic arm
The smart robot growth trend is forecast to be at 80% over the next three years.

What's Next Magazine

Smart robots are on a huge growth trajectory, driven by their potential to boost manufacturing efficiency by handling routine tasks and advancements in enabling technologies for edge AI.

How big is the smart robot trend? To put things into perspective, the CAGR of this segment of industrial robotics is projected to be at 80% over the next three years. The gains in efficiency notwithstanding, it is the development of the enabling technologies including smart sensors, IoT networks, embedded intelligent devices and crucially, advanced processors unlocking the power of edge AI, that that will propel the rapid development and deployment of AI-powered robots.

The market for industrial robotics

Projected to be one of the fastest-growing segments of industrial robotics, AI-powered robots are set to transform the manufacturing and logistics sectors, as well as healthcare, transportation, security and retail applications. It’s not a market to be missed.

 
Market predictions show significant worldwide growth, from $17 billion in 2024 to $30 billion by 2027. By 2032 the market is projected to reach $82.5 billion, propelled in part by the increasing availability of enabling technologies. (Source: Statista)

What does AI bring to robotics?

Robots in the form of programmable machines for automating repetitive tasks have been part of the factory landscape since the 1960s. Initially, robots were only able to respond according to their programming inputs. Later, a limited amount of machine learning and pattern recognition techniques allowed machines to adapt to minor variations in the tasks they carried out. But until now, a robot primarily has been dedicated to recurring, repetitive tasks covering a specific set of circumstances.

What AI brings to the next generation is the ability to adapt to dynamic environments, interact with people and execute more complex tasks. AI neural networks enable robots to respond appropriately to more general situations rather than specific instances. AI-powered robots can recognize and process data from a more diverse range of sources, including sensors and embedded devices. And importantly, they can reference previous experiences, which eventually, will enable them to learn contextually, reason and make decisions.

Artificial neural networks (ANNs) and natural language processing (NLP) are enabling more generalized machine learning. While ANNs attempt to mimic the workings of the human brain, NLP is involved with the interpretation and use of human language by computers. Generative AI builds on NLP, with the iterative learning of large language models to create new content.

The industrial AI robotics market can be broken down into two components. AI industrial robots are for use in manufacturing plants, warehouses and logistics operations. AI service robots are aimed at consumer-facing applications, such as automotive, retail and healthcare.

Making connected devices possible

Even without adding artificial intelligence or machine learning, robots are smart. A lot of the intelligence still relies on the cloud. And more often, cloud-connected robots are operating right alongside human workers. This is putting more focus on safety, security and functionality. Integrating these features only increases design complexity.

AWS Services, including AWS IoT Core, AWS Greengrass, and Amazon SageMaker Edge, make connected devices possible. To make the integration of these enabling AWS Services simple, Avnet created /IOTCONNECT™. Avnet has worked with AWS to abstract out the low-level interfaces between services. /IOTCONNECT is making it simpler to develop and deploy connected solutions like autonomous robots at scale.

To get started with /IOTCONNECT, explore the many development boards available now, or contact the /IOTCONNECT team to discover how Avnet and AWS can help you accelerate your IoT design.

 
 
 

Growing processing power enables AI adoption

Significant increases in compute power, particularly in the form of highly parallel processors, are enabling not only more sophisticated neural networks and AI algorithms but also powerfully effective distributed computing to deliver AI on the edge. This, together with the wider deployment of the Internet of Things (IoT), is opening a potentially huge market.

With today’s AI technology, companies are better able to train machines successfully with vastly more complex models. The latest AI algorithms can understand voice and text (language), images including faces, physical parameters such as temperature, sound, movement, smells, and a host of other analog, unstructured information.

Cue IoT sensor networks, which have been widely deployed throughout factories in recent years. A vast range of ever more sophisticated industrial sensors, radar and cameras can be easily combined and interconnected with complex microprocessors, graphics processors and image processing devices and subsystems. The resulting collection of masses of diverse data provides the fuel required to generate more accurate AI models. More accurate models mean more intelligent operation.

The role of edge AI

Edge computing enables operation in real time, eliminating the delay previously encountered by uploading data to the cloud for analysis. This allows AI-powered machines to show their full potential and better mimic human cognition.

Indeed, edge AI brings significant advantages. In a production environment, a machine might need to analyze historic data or hugely complex models to perform what might be a similar task in different circumstances. This could not previously be achieved through cloud computing due to latency or bandwidth issues, and potentially, security risks. Increasing deployment of 5G networks is further boosting IoT and edge computing capabilities, with faster, more stable and more secure connectivity.

AI at the edge cuts latency, enabling robots to respond faster to external stimuli.

Cloud computing will still be critical for success: For research and longer-term development of AI algorithms, for collaboration and sharing data across global manufacturing sites, and it will continue to support edge AI deployment.

The driving forces of automation: constant change

The driving forces behind today’s fast adoption of AI-powered robots are much the same as they have been for decades. The need to compete globally and on a timely basis means keeping the cost of production down while improving productivity and efficiency. Investment in AI-powered automation is delivering all these benefits while also improving product quality, safety and security.

In a factory or warehouse, machines can accomplish more complex tasks faster, often without or with less human interaction. Edge computing can reduce networking costs while enhancing data security. Further benefits accrue when data can be collected and automatically analyzed for preventive maintenance schedules.

There are both micro- and macro-economic factors further influencing the drive to increased automation. For example, the aging demographics and resulting reduced workforce of some advanced nations require greater productivity.

In the consumer world, we have become used to systems that remember our preferences and/or provide virtual assistants and fast, secure automated ordering and payment. For the innovators, especially in the healthcare sector, AI-powered robots will stimulate the development of new products, processes and services.

For the future, AI developers are researching sequential intelligence, deep learning, supervised and unsupervised learning, reinforcement learning, generative AI and generating emotional intelligence. How these might benefit the industrial world is speculative at present, but if achievable will certainly impact most other aspects of human life. Curiously, some note that it is easier to develop an AI-powered robot to win persistently at chess, for example, than to work out how to pick up an object it has not seen before.

Examples of smart robot technology using edge AI introduced in 2024

 
chart
 

The bottom line

Today, OEMs in the industrial automation sector should embrace the opportunities arising from AI-powered robotics. A wide range of embedded devices, subsystems and platforms for developing AI edge and AI robotics applications are becoming available. This market is on the march.

 

 

About Author

generic graphic of a headshot
Karen Field, Vice President, Content Marketing

Karen Field has spent the majority of her career creating content for electronics design engineers, ...

Additional Information

Why artificial intelligence at the edge matters

How to design a cobot

AI robots worldwide (Statista)

Fortune Business Insights

Helpful Links

Marketing Content Spots
Related Articles
Related Articles
virtual reality headset
Beyond the datasheet: The no-code route to MEMS-based machine learning at the edge
By Philip Ling   -   January 17, 2025
New MEMS sensors from STMicroelectronics integrate an innovative machine learning core, making it simpler to deploy machine language in many applications where motion detection is used. We take a look beyond the datasheet to see how it works.
globe
2025 supply chain insights you won’t get from a genAI chatbot
By David Paulson   -   January 16, 2025
What are the biggest risks and/or opportunities facing stakeholders across the high-tech supply chain in 2025? The greatest opportunities lie in the details others overlook.
Related Events
Related Events