Jan Petrš co-founded Déploya Robotics, an EPFL spin-off that is developing deployable robotic systems for automated façade installation. He discussed that work, and his wider research on novel robot designs for construction and architecture, in Episode 167 of the Robot Talk podcast, published on 23 October 2026. Host Claire spoke with him about how construction robotics might move beyond conventional machine designs.
Who Petrš is
Petrš describes his work as sitting at the intersection of robotics and construction. He is a roboticist, an architect and an entrepreneur. That mix of training shapes the problems he chooses. Architecture supplies the question of what gets built and how it is assembled. Robotics supplies the question of what machine could do the assembling.
He is affiliated with École Polytechnique Fédérale de Lausanne (EPFL) and with Déploya Robotics. Before founding the company, he was a postdoctoral researcher at EPFL's Laboratory of Intelligent Systems. His focus there was tensegrity robotics.
His doctoral research was carried out at the Czech Technical University. It centred on a robotic system that combined reconfigurable robotics with soft robotics for architectural and construction applications. The two fields rarely appear together. Reconfigurable robots change their own shape or structure to suit a task. Soft robots rely on compliant, deformable bodies. Petrš's thesis work aimed to use both in building-related settings.
Déploya Robotics and façade installation
Déploya Robotics is an EPFL spin-off. Its stated goal is deployable robotic systems for automated façade installation. A façade is the outer envelope of a building, and installing one is a defined task within a larger construction project. Picking a single task like this gives a robot a clear job to do.
The word deployable matters here. It points to systems designed to be brought to a site and put to work, rather than machines that only operate in a controlled laboratory. The source does not say how Déploya's systems are built, what they lift or how far along the company is. It does not name customers, funding, pricing or a product. Those details were not part of the published description of the episode, and this report does not supply them.
From tensegrity to the building site
Petrš's postdoctoral subject, tensegrity robotics, is a distinct approach to robot structure. The source gives no technical detail on how he applied it. It does establish the order of his career: soft and reconfigurable systems for architecture during his PhD, tensegrity robots at EPFL, and then a company aimed at façade installation.
That path shows a researcher moving from unconventional robot bodies toward a specific commercial task. The episode is framed around novel robot designs for construction and architecture, which suggests the body of the machine is central to his thinking, not only its software.
The episode in context
Robot Talk is a weekly podcast about robotics, artificial intelligence and autonomous machines. Episode 167 followed a conversation with Rika Antonova of the University of Cambridge about using artificial intelligence to develop complementary robot hardware and algorithms. Before that, Claire spoke with Kushant Patel of Brookhaven National Laboratory about robots that work in scientific facilities such as particle accelerators. Each episode takes a different corner of the field, and this one covers building.
Analysis: why this matters and what is unproven
The most useful signal in Petrš's profile is the narrowness of the target. Automated façade installation is one task on a building project. Construction robotics has often been discussed in terms of general-purpose machines that could do many jobs. A company built around one defined installation task has a clearer case to make. It can be judged on whether it does that task safely and reliably on a real site.
The link to his research is also worth noting. His earlier work used reconfigurable, soft and tensegrity designs, which are less common in the sector than standard arms and mobile platforms. The source does not say which of those ideas, if any, are used in Déploya's systems. A reader should not assume that tensegrity or soft robotics appear in the company's product. The source only establishes that the founder has worked in those areas.
Several questions remain open, and the source answers none of them. It gives no performance figures, no site trials, no payload or reach, and no timeline for deployment. It does not say whether the systems are in use. It does not compare Déploya's approach with other façade installation methods, such as cranes or manually operated lifting equipment. Without that evidence, the claim that novel designs suit construction work is a research direction, not a demonstrated result.
What the episode does show is that academic robotics groups are producing spin-offs aimed at specific building tasks, and that the people behind them often have dual training in architecture and engineering. That combination is a reasonable fit for a sector where the robot has to work with the design of the building as well as the tools on the site. Whether the approach holds up will depend on data Déploya has yet to publish, at least in the material available here. The conversation with Claire is a starting point for that scrutiny, not a substitute for it.