At this year’s opening ceremony, the two panel discussions focused on AI and its impact and importance for the rail sector.
The rolling stock manufacturers were represented by Henri Poupart-Lafarge, CEO of Alstom, Javier Martínez Ojinaga, CEO of CAF, and Michael Peter, CEO of Siemens Mobility.
InnoTrans 2024: From Hype to Reality – AI in the Mobility Sector (left to right: Henri Poupart-Lafarge, Javier Martínez Ojinaga, Michael Peter)
In terms of AI’s entry into the rail sector, Mr Poupart-Lafarge said rail was predestined for this technology because of the vast volumes of data that the industry generates. He did however want to draw a distinction between two phases. The first was data science and machine learning, which he said had been used in the industry for some time now, for things such as predictive maintenance. The second was large language models – LLMs – generative AI and this phase was new. How this area would affect the rail industry remained to be seen, but Mr Poupart-Lafarge was clear that it would.
Mr Peter said AI was as important as the computer and the internet – a step change in how the world works. He said that companies that failed to start using this technology would fall behind and into oblivion in the way as companies that had not adopted computers.
Mr Martínez Ojinaga seemed the most cautious about AI. He said the industry would need to find the right use cases and cut through the hype so that it could deliver substantial value.
Energy Usage
One reason for this, he said was that the rail industry was a climate friendly industry and AI used huge amounts of energy. Therefore we should be intentional with its deployment. He stressed that the most direct benefit of AI was increasing the energy efficiency of train operations, for example through delivering more efficient train driving. These benefits should not be squandered through upping energy usage thoughtlessly – the decision was a trade-off, he concluded.
Mr Poupart-Lafarge did concur that this technology used a large amount of energy. However, he was very clear that the gain of using artificial intelligence far outweighed the problems caused by that and could be mitigated by using renewables to power it.
One example was that through the ability to analyse the extremely large data volumes generated by the rail industry and by adding additional parameters, e.g. in traffic management, it could become possible to filter by or optimise for energy efficiency.
Currently the rail system was far from optimised, he said. The number of trains per track and the number of passengers per train could be drastically improved.
This sentiment was echoed by Mr Peter, who pointed out that once a train was running, its costs were relatively fixed, so if only 40% of seats were occupied, that was hugely inefficient. This is something AI could definitely help resolve.
Recruitment
Over the years the industry has been quite worried about its future workforce. A large number of people are set to retire without a new cohort coming in to replace them. Digitalisation has often been mentioned as a helping hand with this problem. Digitalisation isn’t taking jobs away from people – it is filling roles that are vacant. The same reasoning has been highlighted when the rail industry has tried to diversify its workforce – appealing to wider groups of society than those conventionally working in the industry.
However, with the emergence of AI, the sense among the speakers was far more positive. Artificial intelligence was an exciting, modern field to work in and a field that young people were excited to work in. Coupled with the upcoming generation’s increased awareness about climate change, rail was an industry that allowed them to apply their skills in this area in a sector that was meaningfully aligned with their values.
Mr Peter said that up until now recruitment had focused heavily on mechanical engineers. Now, with AI, science engineers and more mathematicians were also needed – widening the kind of applicant the rail industry might appeal to. In this regard, he called AI a game-changer.
He added that it could also contribute to keeping the workforce happier. One of the main complaints in his company was the number of processes. Of course, because of the critical nature of the work, the business was naturally process-heavy, but much of this could be tackled by AI.
Mr Poupart-Lafarge raised one concern in this regard. He said that if AI replaced a lot of low-value jobs or entry level jobs, then the industry needed to make sure it could train new entrants to the workforce in the long-term. If AI did all the basic coding, efforts must be made to ensure there was no knowledge loss.
The European Picture
One of the challenges faced by the industry in terms of artificial intelligence was the legal framework challenge, Mr Martínez Ojinaga said.
Mr Poupart-Lafarge also stressed that compared to China and the United States, R&D into artificial intelligence in Europe was a dwarf. What was needed was for Europe to come together, instead of remaining scattered, because for AI to succeed, it had to succeed at the European level.
Comment
by Josephine Cordero Sapién
Published
24 Sep 2024
Tags
Alstom
Artificial Intelligence
CAF
Climate Change
Digitalisation
InnoTrans 2024
Siemens Mobility
At this year’s opening ceremony, the two panel discussions focused on AI and its impact and importance for the rail sector.
The rolling stock manufacturers were represented by Henri Poupart-Lafarge, CEO of Alstom, Javier Martínez Ojinaga, CEO of CAF, and Michael Peter, CEO of Siemens Mobility.
In terms of AI’s entry into the rail sector, Mr Poupart-Lafarge said rail was predestined for this technology because of the vast volumes of data that the industry generates. He did however want to draw a distinction between two phases. The first was data science and machine learning, which he said had been used in the industry for some time now, for things such as predictive maintenance. The second was large language models – LLMs – generative AI and this phase was new. How this area would affect the rail industry remained to be seen, but Mr Poupart-Lafarge was clear that it would.
Mr Peter said AI was as important as the computer and the internet – a step change in how the world works. He said that companies that failed to start using this technology would fall behind and into oblivion in the way as companies that had not adopted computers.
Mr Martínez Ojinaga seemed the most cautious about AI. He said the industry would need to find the right use cases and cut through the hype so that it could deliver substantial value.
Energy Usage
One reason for this, he said was that the rail industry was a climate friendly industry and AI used huge amounts of energy. Therefore we should be intentional with its deployment. He stressed that the most direct benefit of AI was increasing the energy efficiency of train operations, for example through delivering more efficient train driving. These benefits should not be squandered through upping energy usage thoughtlessly – the decision was a trade-off, he concluded.
Mr Poupart-Lafarge did concur that this technology used a large amount of energy. However, he was very clear that the gain of using artificial intelligence far outweighed the problems caused by that and could be mitigated by using renewables to power it.
One example was that through the ability to analyse the extremely large data volumes generated by the rail industry and by adding additional parameters, e.g. in traffic management, it could become possible to filter by or optimise for energy efficiency.
Currently the rail system was far from optimised, he said. The number of trains per track and the number of passengers per train could be drastically improved.
This sentiment was echoed by Mr Peter, who pointed out that once a train was running, its costs were relatively fixed, so if only 40% of seats were occupied, that was hugely inefficient. This is something AI could definitely help resolve.
Recruitment
Over the years the industry has been quite worried about its future workforce. A large number of people are set to retire without a new cohort coming in to replace them. Digitalisation has often been mentioned as a helping hand with this problem. Digitalisation isn’t taking jobs away from people – it is filling roles that are vacant. The same reasoning has been highlighted when the rail industry has tried to diversify its workforce – appealing to wider groups of society than those conventionally working in the industry.
However, with the emergence of AI, the sense among the speakers was far more positive. Artificial intelligence was an exciting, modern field to work in and a field that young people were excited to work in. Coupled with the upcoming generation’s increased awareness about climate change, rail was an industry that allowed them to apply their skills in this area in a sector that was meaningfully aligned with their values.
Mr Peter said that up until now recruitment had focused heavily on mechanical engineers. Now, with AI, science engineers and more mathematicians were also needed – widening the kind of applicant the rail industry might appeal to. In this regard, he called AI a game-changer.
He added that it could also contribute to keeping the workforce happier. One of the main complaints in his company was the number of processes. Of course, because of the critical nature of the work, the business was naturally process-heavy, but much of this could be tackled by AI.
Mr Poupart-Lafarge raised one concern in this regard. He said that if AI replaced a lot of low-value jobs or entry level jobs, then the industry needed to make sure it could train new entrants to the workforce in the long-term. If AI did all the basic coding, efforts must be made to ensure there was no knowledge loss.
The European Picture
One of the challenges faced by the industry in terms of artificial intelligence was the legal framework challenge, Mr Martínez Ojinaga said.
Mr Poupart-Lafarge also stressed that compared to China and the United States, R&D into artificial intelligence in Europe was a dwarf. What was needed was for Europe to come together, instead of remaining scattered, because for AI to succeed, it had to succeed at the European level.