A team of roboticists from Ecole Polytechnique Fédérale de Lausanne and economists from the University of Lausanne have developed a new method to calculate which existing jobs are more at risk of being automated away by machines in the near future.
The study was published in Science Robotics.
The team also developed a method to suggest career transitions to jobs less likely to be automated and with the smallest retraining efforts.
Prof. Dario Floreano is Director of EPFL’s Laboratory of Intelligent Systems and lead author of the study.
“There are several studies predicting how many jobs will be automated by robots, but they all focus on software robots, such as speech and image recognition, financial robo-advisers, chatbots, and so forth,” Prof. Floreano says. “Furthermore, those predictions wildly oscillate depending on how job requirements and software abilities are assessed. Here, we consider not only artificial intelligence software, but also very intelligent robots that perform physical work and we developed a method for a systematic comparison of human and robotic abilities used in hundreds of jobs.”
Developing the Method
The team was able to map robot capabilities on job requirements, which was the major breakthrough of the study. They looked at the European H2020 Robotic Multi-Annual Roadmap (MAR), which is a strategy document by the European Commission that is periodically revised by robotics experts. The MAR details which abilities are required from current robots or may be required by future ones. These are organized into categories like manipulation, perception, and interaction with humans.
The team analyzed many research papers, patents, and descriptions of robotic products to assess the maturity level of robotic abilities. They relied on “technology readiness level” (TRL), which is a scale for measuring the level of technology development.
When it came to human abilities, the researchers used the O*net database, which is a widely-used resource database on the US job market. It classifies around 1,000 occupations while detailing the skills and knowledge needed for each.
The team first selectively matched the human abilities from O*net list to robotic abilities from the MAR document, which enabled them to calculate how likely each existing job is to be performed by a robot in the future. If a robot is good at a job, the TRL is higher.
Ranking the Jobs
After carrying out this analysis, the result was a ranking of 1,000 jobs. One of the lowest on the list was “Physicists,” while “Meat Packers” was one of the highest. Jobs in food processing, building and maintenance, and construction had the highest risk.
Prof. Rafael Lalive co-led the study at the University of Lausanne.
“The key challenge for society today is how to become resilient against automation,” Prof. Lalive says. “Our work provides detailed career advice for workers who face high risks of automation, which allows them to take on more secure jobs while reusing many of the skills acquired on the old job. Through this advice, governments can support society in becoming more resilient against automation.”
The authors created a method to find any given job an alternative job with a significantly lower automation risk. These jobs were also close to the original one when it came to abilities and knowledge required, which helps keep retraining efforts to the minimum.
This new method could be used in many different ways. For one, governments can use it to measure how many workers could face automation in the future. This would help tailor retraining initiatives and policies accordingly. Companies could also use it to analyze the costs associated with automation.
All of this work was translated into an algorithm that can predict the risk of automation for hundreds of jobs while also suggesting career transitions.
You can find the publicly accessible algorithm here.
- New Tool Improves Robot Grippers for Manufacturing
- Researchers Develop Amphibious Artificial Vision System
- How to Hire the Best AI & Machine Learning Consultants (2022)
- Is DALL-E 2 Just ‘Gluing Things Together’ Without Understanding Their Relationships?
- Diagnosing Mental Health Disorders Through AI Facial Expression Evaluation