How shall we design robots with human in mind?
In this article, I will explore the subject of robotics, specifically, robotics within cities. I will present four thematic areas for further exploration as well as identifying some of the obstacles that must be overcome for the advances in robotics to impact upon how we think about cities in the future.
In reviewing related research into the robotics field, it is easy to get excited by the sophisticated robotic technologies that already exist. Under the careful supervision of scientist, they may run, play games or move in unison as a swarm of birds. However, when we imagine the same robots outside of their labs and artificial test environments, we picture them navigating real streets, with bumpy roads, unpredictable humans and temperamental weather. This is a very different scenario, where robots are interacting with very unsophisticated technologies (inconsistent internet coverage, limited access to power supply, etc.) and this inevitably surfaces a multitude of concerns and challenges around the technical, social and design requirements for city-based robots.
In this article, I will share my thoughts about ‘robots in cities’ and envision how we might interact with robots, or even how they might interact with us.
What is a robot?
Firstly, I would like to establish a definition of what a robot is. Definitions significantly vary depending on the industry. However, definitions are very important as robots start to appear in our cities because how we define robots could influence social attitudes towards them. For example, the European Parliament has been leading discussions recently about categorising robots as ‘an electronic person’, so they could be taxed in the same manner as a working individual.
That said, defining what a robot is does not present a simple task, especially in today’s information driven world, where digitalisation is affecting so many industries and settings. For example; would we identify an ATM, ticket barriers or dynamic displays as a robotic entity? Or how about bodyless-bots, that are populating service industries, would they deserve to be called a robot?
For the purpose of this article I propose that two attributes — mechanics and intelligence — are essential in order to be categorised as a robotic entity. The level of intelligence could differ: from a simple piece of code with a set of prescribed tasks on-loop, to an infinitely more complex decision-making system, that could process a multitude of information channels from its own sensors and external data inputs simultaneously. In this scenario the robot may even be learning from its own (and ideally other robots or AI’s) actions.
The second essential attribute is mechanics. Again, as with intelligence, robots may be equipped different levels of mechanical abilities; from simple moving parts, like blinking eyelids (often seen on highly stylised ‘cute’ robot designs), through to a highly complex exoskeleton, that is built of hundreds of actuators, motors, microcontrollers and other gears.
When we consider a robot’s physical appearance we find that there is no single design aesthetic or type. Robots can look like anything; a cylinder, a sphere, a foldable origami, a vehicle, it can be small like an insect, or human size, it can fly or crawl, it can be shape-shifting, inflatable, it can be rigid, or it can be soft and flexible like an octopus or it could even be a node living in a distributed intelligence of swarm robotics.
Robotics in the city
Keeping in mind the definition outlined in the previous paragraph and the myriad of potential forms that a robotic entity might manifest itself, we can assume that there are plenty of possibilities for where robotics can be introduced in the city environment. From the perspective of a designer this is an incredibly interesting challenge to consider. There are so many elements to design, starting from the interface of the robot itself, it’s communication with a person or a group of people (community). It’s fascinating to think of what can it do, it’s functions. And no less significant, the design of the city itself — a discussion that surfaces the question of whether our cities, in their current state, are suitable or ready to host this diverse bunch of peculiar city-bots. Would roads, markings, pavements, building’s surfaces, street lamps, shop facades, signage, have to be re-designed? How should existing city infrastructure be modified to suit the needs of robots, in similar way to electric vehicles now requiring electric charging points at regular intervals. In this article I will play with these questions through a set of hypothetical scenarios where ‘electronic persons’ are seamlessly integrated into our city experiences.
In my research I have been particularly intrigued by the idea of hybrid robots. As with many other research fields, research in robotics is heavily dependant on industry funding. Therefore, the evolution of robots for commercial use may become disproportional to those that are assistive to individuals or specific communities.
Right now, you can observe mini-autonomous delivery vehicles on what appears to be unassisted journeys around some areas of London. This scenario, of a goods delivery robot is relatively easy to imagine. But what if the role of this robot wasn’t just delivery, what if it had other functions that when combined would benefit the areas where it is performing these deliveries. For example, the hybrid robot could be programmed to create it’s route through as many green spaces or across grass verges as possible, where it would plant grass seeds on the way and improve the local environment.
Other examples of hybrid robots could include a child’s toy robot that combined with an environmental sensing capability. The toy hybrid robots would be small, they can fly and children would take them everywhere. And because there are many of them, these toys will be acting like a swarm of robots, performing distributed environmental sensing.
Night time robot economy
When we imagine robotics within a messy city environment, we tend to imagine a single robot moving around the city. But more likely it will be dozens, or hundreds of the same or different types of robots across the city. Our cities are crowded already — people, machines, building — density is very high and will only increase. How can we possibly accommodate these creatures? It is easy to picture irritated pedestrians stumbling upon a small robot, hating their slowness and impolite unresponsiveness. One quick solution might be that we make them say sorry with every inconvenience that they cause. Another way to avoid uncomfortable overcrowding could be to distinct the time when some kinds of robots, like maintenance, repair, construction and cleaning are out on the streets for their duties. Streets tend to be emptier, the robots are powered by sun energy that been accumulated during the day and their night vision camera, sonar and other sensors would work perfectly at night. Their communication between each other is silent, their moves are coordinated and undisturbed.
Community robots. Interfaces
But maybe we don’t want to completely remove robots from our city experiences; robots are fun, and could be great to engage with. What if robots could somehow be useful to the community?
Let’s imagine an urban farming robot. It looks after undesirable pieces of land, such as dry ground along main roads or around trees. Or how about imagining another kind of robot that is responsible for filling potholes. This robot would know where he is the most needed to do their job? It could even be connected to an app such as FixMyStreet and pick up tasks from there. Or maybe there is a way that people could book this ‘FixRobot’ directly? Maybe a robot such as this should have an interactive interface, so local residents could book ‘little gardener’ to assist them with their own land. It will scan their garden first, through this interface residence could map/mark/draw where they would like flowers to be planted. They would feed the robot with seeds and let them do the rest. Alternatively they could select an option of what you would like garden to be, for example; wild flowers that are good for bees population, or a garden with only blue flowers, or perhaps just a fruit and vegetable patch? Would we like to programme our robot’s tasks or would we allow a level of creative freedom?
All of the scenarios explored within this article shows that robots in cities won’t all be completely autonomous — urban robots would still need some form of human assistance. Maybe a group of farming and maintenance robots would be brought to specific locations in another vehicle. After being unloaded, they might need to be monitored by a human whilst the task is being performed to make sure that all is fine. In fact, it may even be beneficial for a harmonious human and robot relationship to exist as robots could ask for human help; for example, asking for a lift to the nearest charging point when the robot is low on power. Those dependencies can provoke empathy, appreciation and a sense of responsibility that would be important if robot designers and manufacturers want to ensure acceptance of their technology by society.
The aim of this article is to contribute to the existing discussions on urban robotics from a design perspective. Supported with the examples above, I suggest that if designed properly, robots in the city needn’t be a source of fears, but rather a beneficial thing for the community and individuals.
Few questions have been raised about possible design interfaces, should robots be fully autonomous? what level of decision making should a robot have? These and many more other questions have yet to be explored. I hope there will be more discussions initiated across different fields of design, engineering, urbanism and city planning, to re-assure that technologies that might populate our cities in the future truly consider citizens and are built for social advantage.
Interested in urban robotics?
If you like to learn more or share your ideas please get in touch by emailing anastasia.vikhornova at network.rca.ac.uk