At the biennial Bosch Connected World expo in Berlin, automotive and robotics professionals discussed how current developments are paving the way toward fully autonomous vehicles and how industrial robots are reshaping manufacturing. Photographs from the event were published by Világi Máté for Telex.
Safety first — user trust is key
According to a Bosch survey cited at the event, 48 percent of people consider safety the most important factor, and 44 percent list it as the primary criterion when buying a new car. Where previous decades focused mainly on physical safety metrics such as crash resistance and component durability, software-assisted systems and in-car artificial intelligence (AI) now play an increasingly central role in buyers' assessments.
Panelists argued that one way to build trust in technology is gradual exposure: as more people encounter and use today’s so-called level-2 driver assistance systems and AI features, societal acceptance may rise. Andreas Carlsson, head of Volvo’s driver assistance systems development, emphasized that manufacturers carry significant responsibility: it is far easier to lose public trust than to win it, so automakers should ensure vehicles work reliably from the moment they leave the showroom.
Speakers acknowledged that some people will never want to ride in autonomous cars—either because they love driving or because they distrust machines. Widespread adoption therefore requires some willingness among the public to take risks and experience the technology; the comparison was made to buses or taxis, where passengers routinely entrust their safety to a driver they do not personally know.
AI "hallucinations" and legal responsibility
Attendees noted that automotive AI systems can suffer from the same kind of failures seen in large language models: they can "hallucinate" or produce incorrect outputs. Such errors could have serious consequences, so manufacturers must eliminate or mitigate these behaviors before self-driving systems become widespread.
The panel also highlighted the need for clear regulation: lawmakers must define who is liable if a fully autonomous vehicle causes an accident—the driver who relied on the system, or the manufacturer that claimed its software was safe. Tesla’s approach was mentioned as an example; in some cases Tesla does not assume responsibility for incidents involving its autonomous features, which may deter people from trying fully self-driving cars.
Humanoid robots, jobs and retraining
The expo covered more than cars: AI, manufacturing processes and robotics were central themes. In his opening remarks, Stefan Hartung, CEO of Bosch, argued that demographic trends—referring to aging populations—make humanoid robots necessary. He added that despite a perceived need, several technological hurdles remain.
Speakers often downplayed the downsides of automation, arguing that while some jobs will disappear, new ones will be created. The author of the report, however, raised practical concerns: how long and how costly is it to retrain someone who has worked for twenty years as a conference interpreter, forklift operator, warehouse worker, copywriter or bus driver? It is unclear who would finance potentially lengthy retraining, which could take years, and which professions will remain secure in the long run.
Why human-like robots?
Exhibitors defended humanoid robots on practical grounds: many industrial environments were built for humans, so humanoid machines can fit into existing spaces and handle objects in ways similar to human workers. Producers also believe people will react more positively to machines with human traits, although the panel did not support this claim with formal studies, relying instead on industry observation.
Conclusion — optimism tempered by caution
The overall impression from the event is that tech companies and industrial players remain strongly optimistic about both near- and long-term prospects, even as they repeatedly cited global supply-chain challenges as potential obstacles. Demonstrations in autonomous driving, big-data processing and industrial AI appeared promising, but safety, reliability, regulation and the handling of social consequences are essential if the technology is to deliver real benefits.
The author attended the event at the expense of Bosch; the company had no editorial influence over the article.
