A new flashpoint in the human-machine interface has emerged on the sidewalks of Milton Keynes, where autonomous delivery robots have been documented issuing auditory commands for pedestrians to “get out of the way”. This behaviour, while efficient in terms of logistics, has ignited a public backlash that underscores the friction between technological optimisation and social norms.
The robots in question, operated by Starship Technologies, have been deployed across several UK towns since 2018. Their purpose is to deliver groceries and parcels within a 4-mile radius, using a constellation of sensors and machine learning to navigate pavements. The command, a pre-recorded message triggered when the robot’s path is obstructed, is designed to maintain a steady delivery tempo. However, residents have expressed frustration at what they perceive as a discourteous imposition. "It feels like being bossed around by a wheelie bin," one local remarked.
This incident is not an isolated one. As autonomous vehicles and robots proliferate, the question of right-of-way in shared spaces becomes increasingly acute. Pedestrians, cyclists, and now robots must negotiate a common ground for which no existing traffic code was designed. The UK government’s Centre for Connected and Autonomous Vehicles estimates that 40% of public interactions with delivery robots result in some form of hesitation or conflict. The core issue is a mismatch of expectations: humans operate on social heuristics of politeness and mutual accommodation, while robots operate on algorithms optimising for minimal delay.
From a technical perspective, the robots’ command is a symptom of a deeper problem: the inability of current AI to read subtle human cues of reluctance or confusion. A robot that senses an obstacle will predict the pedestrian’s next move using a probabilistic model. If the pedestrian deviates from the expected path, the robot defaults to a vocal instruction. This is an engineering shortcut. A more sophisticated approach would involve slowing down, re-routing, or even stopping to allow the human to pass. But such behaviours increase delivery times and reduce battery efficiency.
The backlash, however, is not just about politeness. It touches on a deeper anxiety about the erosion of public domain. Sidewalks have historically been a democratic space, regulated by informal norms. The introduction of a corporate-owned autonomous agent that issues commands overturns that social contract. Residents feel they are being subordinated to the efficiency goals of a private company. This sentiment is amplified by the lack of local consultation. Starship Technologies, while working with local councils, has not always engaged with pedestrian groups.
There are parallels here with the advent of the automobile in the early 20th century, when pedestrians were forced off streets that had once been their shared domain. The difference is the speed of change. Autonomous robots are arriving without the gradual cultural adaptation that accompanied the motorcar. The solution is not to ban the robots, but to mandate a different set of interaction protocols. Cities could require that robots yield to pedestrians in all circumstances, or that they emit no audible commands. Alternatively, the robots could be equipped with more natural language processing to articulate requests rather than demands.
From a climate perspective, delivery robots offer a carbon-reducing alternative to vans. Each robot replaces a short car journey, cutting emissions by up to 80% for that leg. Their adoption is thus aligned with net-zero targets. But the public backlash threatens to slow deployment, which would be a setback for urban decarbonisation. Engineers must therefore treat this as a user interface problem. The robot’s social interface needs to be redesigned to engender trust and compliance rather than resentment.
In the end, the “get out of the way” directive is a canary in the coal mine. It signals that we are automating not just tasks but relationships. How we design these interactions will shape the liveability of future cities. The robots are listening. It is time we taught them manners.
