The Robot That Solved the Last Hundred Metres
Last-mile delivery, the final segment of the parcel or food delivery journey from a local depot or restaurant to the customer's door, is the most expensive and labour-intensive component of the delivery logistics chain, accounting for approximately fifty percent of total delivery cost despite covering only a small fraction of the total distance the shipment travels. The cost concentration in last-mile delivery reflects the economics of urban delivery density whose scattered delivery destinations, traffic-constrained access, and the labour cost of the human courier or driver whose time is consumed by navigating to each address and waiting for the customer to collect the delivery creates the unit economics that have made on-demand food and parcel delivery commercially marginal despite the enormous consumer demand that the gig economy delivery platforms have demonstrated. Sidewalk delivery robots address the last-mile cost problem by replacing the human courier for short-distance deliveries whose weight, time sensitivity, and geographic proximity to the robot's operating area make autonomous ground-based delivery operationally feasible at a cost below the human equivalent once the robot's capital cost is amortised across the deliveries it completes over its operational life.
The autonomous delivery robot market, valued at approximately $1.33 billion in 2026 and growing at nearly twenty percent annually toward $3.27 billion by 2031, reflects the transition from the pilot programme and technology demonstration phase that characterised the industry through 2022 to the operational scaling phase that the deployment data of 2024, 2025, and 2026 is confirming. The industry's commercial progress is most accurately measured not by capital raised or technology announcements but by deliveries completed per robot per day and by the route completion rate without human remote assistance intervention, both of which are improving as the machine learning systems underlying the autonomous navigation mature on the operational data that fleet-scale deployment generates. The regulatory framework for sidewalk robots, which requires state-level legislation in the United States specifying the permissible operating parameters including maximum speed, weight, and geographic restrictions that sidewalk robots must comply with, has been enacted in over forty US states and is expanding internationally as European and Asian cities develop their own frameworks for autonomous ground vehicle operations on pedestrian infrastructure.
Starship's Ten Million Delivery Milestone
Starship Technologies' announcement in April 2026 that its fleet had completed more than ten million deliveries represents the most commercially significant operational milestone in the autonomous delivery industry and the clearest evidence that sidewalk delivery robots have crossed the threshold from technology demonstration to genuine commercial service. Starship's network of more than three thousand autonomous robots operates across more than three hundred locations in eight countries, has travelled over twenty-two million kilometres autonomously, and has recorded approximately 1.8 million kilograms of avoided carbon dioxide emissions whose environmental impact the company reports as a regulatory and marketing asset in European cities tightening urban freight sustainability requirements. The company's campus-to-community expansion strategy, which begins with university campus deployments where the defined geographic boundaries, predictable demand patterns, and sympathetic regulatory environment create ideal operating conditions and then extends into the suburban residential and mixed-use commercial environments whose delivery density and pedestrian infrastructure characteristics differ from the campus context, has created the geographic scaling model that its competitors are adapting for their own market development strategies.
Serve Robotics, spun out of Uber's Postmates subsidiary in 2021 and having gone public on Nasdaq, deployed one thousand third-generation autonomous sidewalk delivery robots in 2025 and established itself as the primary Uber Eats delivery robot partner in Los Angeles, Miami, Dallas, Chicago, and Atlanta, reporting more than one hundred thousand completed deliveries from its LA market alone. Its third-generation robot's improved sensor suite and autonomy stack, which reduces the remote human assistance interventions that earlier generations required for the complex urban pedestrian environments that Los Angeles deployment encounters, demonstrates the technology maturation that operational scale enables through the machine learning improvements that larger deployment generates. Coco Robotics, whose remotely supervised model employs human operators who can intervene when the robot's AI encounters edge cases it cannot handle autonomously, has set the most ambitious fleet target in the industry at ten thousand robots by the end of 2026, with its human oversight model allowing faster deployment in complex urban environments whose regulatory and operational challenges autonomous-only systems cannot yet handle without intervention.
Unit Economics and the Commercial Viability Question
The commercial viability of sidewalk delivery robots depends on the delivery economics whose unit cost must compare favourably with the human delivery cost for the specific delivery types, order values, and geographic densities that the robots serve. The grocery and prepared food delivery applications that Starship, Serve, and Coco primarily serve create the high delivery frequency per robot per day and the consumer willingness to pay delivery fees that make the economics most favourable, because frequent same-day deliveries of relatively small orders whose customer is typically at the delivery address create the conditions where robot delivery competes most directly with human couriers rather than with parcel delivery whose longer delivery windows and larger package sizes create different operational parameters. The estimated delivery cost of approximately one dollar per delivery that Starship targets at operational scale compares with the approximately five to eight dollar cost of human courier delivery in the same markets, creating the economic case that justifies robot fleet investment at the delivery volumes that the company's operational scale is approaching in its most mature markets.
Top 10 Companies in Autonomous Sidewalk and Last-Mile Delivery Robots Globally
- Starship Technologies: Estonian-US sidewalk delivery robot company with 10 million completed deliveries across 3,000 robots in eight countries; its April 2026 milestone and its campus-to-community expansion strategy create the most commercially validated sidewalk robot operation whose delivery volume and geographic breadth define the commercial benchmark.
- Serve Robotics: US sidewalk delivery robot company spun from Uber Postmates with 1,000 third-generation robots and Uber Eats partnership; its Nasdaq listing and its LA, Miami, Dallas, Chicago, and Atlanta deployments create the commercial autonomous delivery robot company whose public market status and food delivery platform integration drive its commercial scaling.
- Coco Robotics: US remotely supervised delivery robot company with 10,000 robot fleet target for end of 2026; its human-supervised autonomy model that scales faster in complex urban environments and its aggressive fleet expansion create the commercial robot delivery company whose supervised autonomy approach trades full autonomy for faster urban deployment.
- Kiwibot: Colombian-US campus and urban delivery robot company with DoorDash and Grubhub partnerships; its colourful robot design and its food delivery platform integration create the campus delivery robot whose university market penetration and food delivery partnership model create the commercial deployment pathway that food delivery demand sustains.
- Nuro: US autonomous vehicle company with street-legal delivery vehicle and pivot to technology licensing model; its $2.3 billion raised and its November 2024 pivot to licensing its autonomous vehicle technology to other companies rather than operating its own fleet create the autonomous delivery technology licensor whose platform serves larger vehicle delivery applications beyond sidewalk robot payload limits.
- Ottonomy: US delivery robot company with Ottobots for campus, airport, and urban delivery; its airport logistics deployment and its multi-environment operating capability create the delivery robot platform for the indoor-to-outdoor transition environments that sidewalk-only robots cannot serve and that airport terminals and mixed-use development require.
- Amazon (Scout heritage): US e-commerce company whose Scout delivery robot programme pioneered the sidewalk robot blueprint before its 2022 cancellation; its continued investment in autonomous delivery including Prime Air delivery drones and its enormous logistics infrastructure create the context that Amazon's future re-entry into ground-based autonomous delivery could leverage through its operational scale and last-mile delivery density advantages.
- FedEx (Roxo): US logistics company with Roxo autonomous delivery robot in limited commercial deployment; its parcel delivery use case for Roxo whose payload and operating parameters differ from food delivery robot applications and its logistics infrastructure create the established carrier's autonomous ground delivery commercial experiment.
- Neolix: Chinese autonomous delivery van company with larger payload vehicles for commercial last-mile logistics; its deployment in Dubai, Beijing, and Europe and its up to 200 kilogram payload capacity create the autonomous delivery vehicle for the heavier parcel and commercial supply delivery applications that sidewalk robots' payload limits exclude.
- Cartken: US delivery robot company with Grubhub and DoorDash partnership for campus and urban food delivery; its robotic autonomy stack and its food delivery platform partnerships create the commercial delivery robot whose operational focus on the food delivery use case creates the narrow application specialisation whose economics are most clearly validated by current robot delivery deployment data.