Industry

Design and human factors make robots usable, says David Robert

David Robert, former head of human-robot interaction at Boston Dynamics, argues that design and human-centered thinking are crucial to making robots acceptable and useful.

Design and human factors make robots usable, says David Robert

David Robert, who for years led human–robot interaction efforts at Boston Dynamics and was responsible for the company’s robots’ personalities and how they connect with people, left the firm this spring and has since started his own project. His new work emphasizes design and human-centered considerations over pure technology, drawing also on traditional craftsmanship; he is currently gathering inspiration in Budapest and collaborating with Hungarian designers and artists, including influences from Hungarian textile art.

From animation to robotics

Robert’s career began in Hollywood, where he worked on animation and special effects, particularly procedural animation — a technically oriented approach that draws on concepts from artificial life. In 2009 he received a scholarship to the MIT Media Lab and joined Cynthia Breazeal’s group, where social and design questions about robots were part of the agenda. He later studied the social science side of robot design at Harvard under Joe Blatt, an expert on children and media.

He also helped launch a robotics division at Huawei and worked at a startup focused on autonomous vehicles. After funding ran out there, he joined Boston Dynamics, supported by the company’s strategic leadership to make room for designers alongside engineers.

Why design matters for robots

Robert stresses that design is more than surface aesthetics: it shapes how a robot behaves. The smoothness and quality of motion, the animation, and behavioral choices affect whether people trust a machine or feel threatened by it. Technological feats alone are insufficient if the resulting robot frightens users; human-centered design and a focus on the user experience are therefore crucial.

Physical limits — motor capabilities and the laws of physics — mean that digital animations cannot always be transferred directly to real robots. Robert cites the evolution of Atlas as an example: the original Atlas used hydraulic systems with safety and reliability issues, while the newest Atlas was redesigned from scratch. He notes Atlas is already in practical use in Hyundai factories and warehouses, illustrating the shift from demo-grade stunts to production tools.

Humanoid form and functionality

One frequent question is why some robots resemble humans. Robert argues the humanoid form is practical in human-built environments — doors, handles, and stairs are designed for people — so a humanoid or human-scale robot can interact effectively with that environment. Boston Dynamics intentionally keeps its robots’ appearance clearly mechanical (for example, a round head) to signal they are tools, not replacements for humans. At the same time, robots like Atlas sometimes move in ways beyond human capabilities, which is acceptable in industrial contexts where efficiency matters.

Where we are and what’s next

Robert places current robotics primarily in industrial settings, with service-sector robots already appearing in parts of Asia and the U.S. (food delivery, hotels, restaurants). Household robotics are still nascent: the robot vacuum is the main example of widespread domestic adoption. He outlines a likely progression: industry → service sector → homes, but emphasizes this will take time and depends on production scale. If factories begin producing hundreds of thousands of units, many robots will inevitably appear in public spaces.

Many small robots vs. one generalist

Robert expects a likely future in which many smaller, task-specific robots operate together under coordination from a central AI or environmental intelligence, rather than a single, universal humanoid robot. Coordinating specialized robots can be simpler and more cost-effective than building one machine that does everything.

Social bonds, emotions and building trust

Humans tend to anthropomorphize machines, projecting emotions onto them; for Robert, building trust is therefore essential. He proposes giving users gradual control: start with a robot performing small, observable tasks so people can see it is reliable, then expand its responsibilities. Studies he references also show families sometimes react to robots’ mistakes as they would to a family pet — correcting them but continuing to bond — though he cautions these are limited-sample findings.

Robert believes consciousness is a spectrum and that, at some point, machines could have experiences that resemble sensations or pain, raising ethical and legal questions. For now, however, he insists robots remain tools serving humans, and that designers should preserve a clear distinction between living beings and machines while thinking ahead ethically.

Regulation and good design in public spaces

He argues for clear standards and regulations so robots do not obstruct people in shared spaces. Practical measures include machine vision to recognize humans and maintain virtual safety boundaries that robots do not violate. Since tens of thousands of robots are expected to be manufactured in coming years, Robert says the industry should insist on physical and psychological safety by design.

The future: coexistence and identity

Robert envisions humans and robots living side by side, possibly becoming intertwined via implants or other integrations, but personally aims to be the best human he can be rather than uploading his mind. He describes robots as a new ontological category — neither dead nor alive, somewhere between — and thinks people will always be able to distinguish robots from living beings. If robots become commonplace in daily life, design, regulation, and social acceptance will need to evolve together.

In his Budapest-based project, Robert focuses on design, social behavior, and human-centric solutions rather than technical spectacle. He argues only by centering human needs and perceptions can robots become truly usable and accepted tools in people’s lives.

Brief chronology (as mentioned in the interview)

  • Career start: Hollywood animation and procedural animation
  • 2009: scholarship to MIT Media Lab; joined Cynthia Breazeal’s group
  • Harvard: studied under Joe Blatt on children and media
  • Later: work at Huawei and an autonomous vehicle startup
  • More than four years at Boston Dynamics; left the company this spring

Photo: Hevesi‑Szabó Lujza / Telex