Angus Muirhead
Thematic Equities

Key takeaways

  • While robotics and automation systems are in use in a number of industries, their role is limited largely to repetitive and predictable tasks that do not require any dynamic interaction with their surroundings.
  • Machine learning and deep neural networks can enable much greater autonomy, opening a wide range of new possibilities for automation systems and robotics.
  • Powered by AI, robotic systems are likely to become significantly more useful to people, businesses, and governments around the world.

As technologies, such as artificial intelligence (AI), continue to advance, robotics and automation systems are likely to become smarter, able to operate with greater autonomy, cheaper and easier to use. As a result, they may become economically viable and useful in many more roles and industries.

What are major applications of robotic automation systems today?

Automation systems are often hidden from view, operating in the background. We often do not notice their presence, except for when something goes wrong, and the system does not work. Although there are a wide variety of use-cases for robotic systems, here are five of the most common:

Automotive industry

Car manufacturing employs more robots than any other industry in the world; welding and gluing parts of the chassis together with great speed and high precision, and typically running almost non-stop for ten years or more. The International Federation of Robotics estimates that car production employs approximately 1 million robots, and Korea leads the way in robot density in the auto industry at approx. 2.9 robots for every 10 factory workers.1 Advances in technology are making new automation solutions available to car makers. A good example are autonomous mobile robots (AMRs). While in a traditional production line vehicles wait in queues, in a modular production system AMRs – mobile platforms – carry entire vehicles to different workstations and sophisticated sensors allow robots to operate safely without physical safety cages. The possibilities in this industry are growing fast.

Semiconductor industry

Robotics and automation are well suited to manufacturing cars, but also to making chips. That’s because cars and chips typically offer ideal production characteristics for automation: large volumes of product, with little variation or “mix”. As the world becomes more digital, semiconductors are being designed into more products than ever before. The process of making semiconductors is one of the most technically sophisticated in the world, typically requiring 11 specific steps,2 many of which are repeated multiple times to produce the chip. The equipment used in all the critical steps is highly automated and specialized, and one of the most sophisticated tools is “EUV lithography”. Dutch manufacturer ASML spent more than EUR 6 billion in R&D to develop their extreme ultraviolet (EUV) tool.3 It sells with a price tag of more than USD 300 million. As the world adopts more AI technology, more automated tools will be needed to produce the chips.


We don’t give a second thought to ordering items online and receiving them a few days later, or even the same day. However, the logistical acrobatics required to make this happen is remarkable. Traditional commerce is much simpler: taking goods from the factory to a distribution hub and then to a regional distribution center and on to a retail outlet. Fixed routes, regular quantities, repeated and predictable. However, e-commerce allows anyone with an internet connection to order from a vast choice of goods, for delivery to any address in the world. Automation and robotics are critical in enabling this to be done fast and cost effectively. Software helps automate inventory management all the way down to individual retail outlets, route planning for shipments, preparation of custom forms and payment of duties, and various robotic solutions facilitate packing, sorting, and handling of the items for shipment. Since e-commerce is expected to grow at 9% a year ,4 the industry needs to invest more in automation to meet this demand.


Not so long ago, industrial designers and engineers would make physical prototypes of new products in development. This was done to uncover unforeseen issues in design and for manufacture, and to test them in real-world scenarios such as submersion in water, extreme temperature and pressure, repeated use, and crash or collision testing. Nowadays, software can automate many of these processes, simulating thousands of real-world scenarios in a virtual world, with great accuracy and faster and more efficiently than could be done with physical prototyping. Since so many simulations can be run, the resulting products are typically more reliable, safer, and long-lasting than in the past. And this area is now evolving fast thanks to some smart AI technologies.

Food processing

Much of our food, fish, meat, vegetables, and fruit, as well as processed food such as pasta, cookies, cakes, and chocolate, are produced with varying degrees of automation. While some types of food present technical challenges to automation due to the wide variety of shapes, sizes, and consistency (there are over 200 varieties of edible fish for example), others are far more uniform and relatively simple to process automatically. The global market for pasta is worth approximately USD 66 billion5 (revenues) and most commercial pasta production is automated. Handmade or “artisanal” pasta may taste nicer, but it is unlikely to be price competitive against the vast scale of the commercial giants. In Parma, Italy, the flagship factory of the world’s largest pasta maker, Barilla Group, is a fully automated “lights-out” operation which runs 24 hours a day 365 days of the year. With 120 laser-guided vehicles and 37 robotics systems, it produces 320 billion grams of pasta every year.6 This type of automation at scale helps feed the world at affordable prices, but it can also minimize waste and reduce the risk of human injuries and food contamination, ensuring food safety standards are met.

Robots are becoming smarter

AI has come a long way since the first digital computers of the 1940s. In the early days, AI were rule-based systems used to make simple predictions and calculations. More recently, machine learning and deep neural networks can enable more creative or “generative” tasks, opening a wide range of new possibilities for automation systems and robotics. At the same time, as technology becomes more widespread tech companies enjoy greater economies of scale, and this enables robotics and automation makers to build systems more cheaply. In turn, lower priced automation solutions attract new customers and find new use cases.

By contrast, human labor is becoming more expensive and people are less inclined to perform certain tasks. In many countries there is a severe shortage of skilled workers in factories, logistics centers, on farms and in hospitals and nursing homes. Businesses are also under increasing pressure to stay competitive and to meet strict regulations to ensure the health and safety of their workforce, and high quality and safe products for their customers. As technology advances, smarter and cheaper automation systems may provide an answer to these challenges.

As robots become smarter and safer to use, they are likely to make their way to a broader range of industries, equipped with skills and capabilities that far surpass those of earlier generations. Here are five of the most significant opportunities for smarter automation in our view:


Approximately 40% of the world’s land is used for farming, one third crops and two thirds’ livestock. As the global population grows and people consume more food, traditional modes of agriculture are leading to unsustainable land degradation. Precision agriculture uses satellite imagery and ground sensors to reduce the use of water, fertilizers, and weed killers. Some solutions claim to cut more than 90% of chemical run-off.7 Smart robotic solutions are also making their way onto farms, autonomously navigating fields to destroy weeds and identify bugs and crop disease. Others are tasked with pruning the rows of crops, harvesting, and sorting and grading fruit and vegetables.8


Recognizing a disease early can in some cases increase the chance of curing it. Medical diagnosis stands to take a huge step forward with the use of AI technologies to find patterns in enormous volumes of medical data (often unstructured data) ranging from human DNA to medical records, family history of disease, blood types, and environmental factors. This promises to greatly facilitate the early prediction of disease, allowing patients time to change their lifestyles and allowing doctors to deliver preventative medicines before the symptoms appear.

Internet of things (IoT)

Singapore has one of the most intelligent traffic systems in the world, with sensors and cameras gathering data in real-time across their network of 160 km of motorways to improve traffic flow and road safety. The government’s “2023 Smart Mobility” plan 9 describes a smart, sustainable, and interactive transport system using GNSS (a.k.a. GPS) data in mobile phones and cars, together with sensors and cameras in the road and traffic light infrastructure, to dynamically change road signs and control traffic lights to improve the flow of traffic and avoid congestion. Over the next decade, as vehicles become more autonomous and intelligently connected, the possibility to improve convenience, traffic flow, and energy efficiency, should increase dramatically.

Generative design

Generative AI technologies are taking product design to new levels of sophistication. Generative design uses AI technologies to produce and evaluate multiple design alternatives, which the engineer can then assess, adapt, and take forward into development. Effectively, it reverses the “simulation” process. The engineer inputs the desired outcome, parameters such as strength, weight, flexibility, physical dimensions, etc, and the system analyzes a vast library of physical properties to create the optimal design. Interesting to note that many of the designs resemble trees, plants, and other elements of our natural world.

Data sharing

Saving possibly the biggest and the best for last, the potential for AI to revolutionize every field of scientific research is huge. So much innovation today is discovered by chance, and the more we understand about specific fields, the more difficult it is for people to attain a polymathic oversight of knowledge across a range of subjects. One example of the great potential opportunity was demonstrated by DeepMind, a UK company acquired by Google in 2014. After beating the world champions at the game “Go”, a type of Chinese checkers, DeepMind turned their attention to predicting the 3D structure of proteins.10 This represents one of the fundamental challenges in biology. DeepMind’s “AlphaFold” team trained AI algorithms on the 170,000 proteins which were publicly known and available at the time, and only a few years later in 2021 published a database of more than 200 million 3D protein structures.11 DeepMind made this vast library freely available to help accelerate scientific research. With giant steps forward in fundamental knowledge of science and our natural world such as these, it is likely that the pace of innovation will experience a period of exponential acceleration.

AI – the driving force behind robotics

Whether we are aware of it or not, robotics and automation are already deeply engrained into many aspects of our lives. Yet as the power of AI and the sophistication of sensors and materials sciences continues to increase, and at the same time, economies of scale and Moore's Law12 continue to lower costs, robotic systems are likely to become significantly more useful to more people, businesses, and governments around the world. The advance of AI and other technologies is one of the key long-term drivers of the robotics investment theme.

Although technology and innovation have already shaped civilisation for thousands of years, we believe we are now entering a golden era of innovation in robotics and automation, and that AI will enable significant steps forward in economic productivity and sustainability.

About the author
  • Angus Muirhead

    Head of Thematic Equities

    Angus Muirhead (BA, CFA), Managing Director, is Head of Equities at Credit Suisse Asset Management, now part of UBS Group, and Lead Portfolio Manager for the Robotics strategy. Angus joined the Thematic Equity team in 2016 as a Senior Portfolio Manager. He started his investment career in 1997 as a buy-side equity analyst at Phillips & Drew Fund Management in London before moving to Tokyo in 2000 to focus on the Japanese technology and healthcare sectors. In 2007, he moved to Zurich as a portfolio manager specializing in global technology and healthcare-related thematic equity funds. Angus holds a bachelor’s degree in Modern Japanese Language and Business Studies from Durham University, United Kingdom, including a year of study at Kumamoto University, Japan, and is a CFA charterholder.

Related insights

Contact us

Make an inquiry

Fill in an inquiry form and leave your details – we’ll be back in touch.

Introducing our leadership team

Meet the members of the team responsible for UBS Asset Management’s strategic direction.

Find our offices

We’re closer than you think, find out here.