In the bustling labs of Carnegie Mellon University (CMU), a quiet revolution is underway. It's not just about creating smarter or faster robots, but about making them more inclusive. In a groundbreaking study, researchers Sarah Fox and Nikolas Martelaro are challenging the status quo in robotics design by doing something surprisingly rare: asking people with disabilities for their input.

The results, presented last month at the prestigious ACM Conference on Human Factors in Computing Systems (CHI 2024), are as enlightening as they are inspiring. By bringing together roboticists and individuals with mobility disabilities, Fox and Martelaro, both assistant professors at CMU's Human-Computer Interaction Institute (HCII), have uncovered not just design flaws in current robots but a wealth of innovative ideas that could redefine our urban landscapes.

The study's genesis lies in a growing concern: the proliferation of sidewalk delivery robots. While these autonomous couriers surged in popularity during the pandemic, their presence isn't always welcome. "They typically compete with pedestrians for limited sidewalk space," explains Howard Han, the HCII Ph.D. student leading the research. For those with mobility disabilities, the consequences can be severe. Han cites a 2019 incident at the University of Pittsburgh where a delivery robot blocked a wheelchair user from accessing a ramp—turning what should be a safe path into an impassable barrier.

Such incidents underscore a critical issue in robotics: accessibility often comes as an afterthought. "Roboticists understand that consideration for accessibility often comes too late in the design process—often at the very end," Fox reveals. But why? The answer lies in the industry's rapid pace. Most sidewalk robot companies are startups, racing to market. "There's not really the time or incentive to slow down," Fox notes, "in the ways needed to get that type of input from a variety of stakeholders, including people with disabilities."

Yet, when given the chance to collaborate, these stakeholders bring invaluable perspectives. Take, for instance, the idea of a grocery-fetching robot. One participant with a disability firmly advocated for a design that could "stand up and reach something on a high shelf." It's a simple yet brilliant concept, born from lived experience—one that might never have occurred to a designer without mobility challenges.

But collaboration also reveals nuanced considerations. When the same participant proposed the shelf-reaching feature, a roboticist cautioned that such a design "could actually scare people because it could be read as creepy or foreboding." This exchange highlights the complex interplay between functionality, user experience, and public perception—factors often overlooked in traditional design processes.

The study's impact extends far beyond refining existing robots. It's sparking ideas for entirely new types of "public service robots." As Fox excitedly shares, "What's fun about the most recent paper is that through this collaborative design opportunity... they conceptualized a number of other ideas for robots that operate on the sidewalk and in public that aren't just for delivery."

Imagine robots that clear sidewalks of debris or snow, ensuring safe passage for all. Picture robotic crossing guards, or AI assistants that carry books and supplies for people, or even guide shoppers through stores. In one particularly innovative concept, robots could act as mobile traffic reporters, detecting blocked paths or obstacles that might impede wheelchair users, then sending alerts via a mobile app—much like Waze does for car traffic.

Yet, as the researchers have documented in previous studies, current sidewalk robots have significant limitations. "Most often, they're operating around campuses," Martelaro explains. "Campuses are more controlled... Also, the population is more likely to accommodate the robots." But in the real world, with its broken sidewalks, parked cars, and unexpected obstacles, many robots get stuck or even fall into planters, requiring human rescue.

Even in their primary function—delivery—these robots often fall short for those with disabilities. While they can transport food from point A to B, many users need items brought to their doorstep or even placed on their countertop. "Robots can't complete that last stretch," Fox notes, "because they don't yet know how to climb stairs or take an elevator."

Addressing such challenges might require dramatic redesigns or entirely new features. As Martelaro puts it, "If you step a little bit beyond a robot that rolls around and has some lights and visuals on it to something that has an arm on it, that becomes a more technically challenging task." But it's precisely these challenges that the CMU team believes must be tackled, not avoided.

Their approach isn't just about making robots more accessible; it's about fundamentally rethinking how robots are designed. By including diverse stakeholders from the outset, they're uncovering use cases and concerns that might otherwise be overlooked until it's too late or too costly to address.

The implications are profound. As urban spaces become increasingly automated, the design choices made today will shape the accessibility and inclusivity of our cities for generations. Will our sidewalks be navigable for all, or will they become obstacle courses for those with disabilities? Will our public robots be universally helpful, or will they cater only to a narrow segment of society?

Fox and Martelaro's work suggests that the answers lie not in the labs of tech startups alone, but in the lived experiences of those who will coexist with these machines. By bringing together roboticists and people with disabilities, they're not just refining technology; they're redefining the relationship between humans and robots in public spaces.

As air travel recovers and cities grow more crowded, the pressure to automate will only intensify. Yet, as this groundbreaking study from Carnegie Mellon shows, speed need not come at the cost of inclusivity. By listening to diverse voices and designing for all, we can create a future where technology doesn't just serve the many, but genuinely serves everyone.

In the end, the most profound insights from this research may not be technical but human. As robotics advances at breakneck speed, it's the simple act of collaboration—of listening and learning from those different from ourselves—that could guide us to a future that's not just automated, but truly accessible.