Siirry sisältöön
Robotised on-demand delivery – keeping humans in-the-loop

Online ordering coupled with fast, on-demand delivery through a distributed courier network has in recent years started to transform how retail products and restaurant food is consumed. Robotised on-demand delivery is set to transform this even further.


Aarni Tuomi

lehtori, majoitus- ja ravitsemisliiketoiminta
lecturer, hospitality business
Haaga-Helia ammattikorkeakoulu


Visiting Research Fellow
University of Surrey

Mário Passos Ascenção

principal lecturer
Haaga-Helia ammattikorkeakoulu

Published : 11.08.2022

In Finland, the on-demand retail and restaurant food delivery market is largely controlled by two platform economy companies: foodora (part of Berlin-based Delivery Hero) and Wolt (part of San Francisco-based DoorDash). Put together, both companies have thousands of active courier partners zipping around Finland’s major cities to complete on-demand delivery tasks. Soon the armada of couriers equipped with scooters, bikes, mopeds and cars (and during Winter, even cross-country skis) might be complemented and to some degree replaced by service robotics.

Despite still not being a thing in Finland, both Delivery Hero and DoorDash, along with other delivery giants, e.g. UberEats and Just Eat Takeaway, have dabbled in autonomous delivery in the past. In 2017 Delivery Hero partnered with Estonian startup Starship Technologies to trial robot-delivery in Hamburg, and in 2021 it partnered with Italian startup Yape to bring robot-delivery to the streets of Stockholm. Wolt’s parent company DoorDash has also partnered with Starship in the past, along with many other robotics companies, and has even gone so far as to establish its own DoorDash Labs with the aim of bringing automation to on-demand delivery.

As part of Haaga-Helia’s AlgoAmmatti-project, we have previously noted how the platform economy, e.g. platforms that facilitate on-demand delivery, enables new ways for social and material actors to entangle in practice. We have also pondered the role of algorithmic management to improve the productivity of the on-demand delivery service process, primarily through optimising the behaviour of delivery couriers.

In this article, we consider how delivery-focused platform economy companies might keep humans in-the-loop by leveraging their existing couriers when bringing automation to on-demand retail and restaurant food delivery.

Keeping humans in-the-loop

In the coming decades, some jobs/tasks will be eliminated, many tasks transformed and new tasks created. Existing couriers have a great deal of domain expertise when it comes to on-demand delivery. As the automation of tasks through service robotics changes the service process (Tuomi & Ascenção, 2021), new roles and tasks are likely to emerge. Existing couriers’ domain expertise would be invaluable and directly applicable across multiple roles. In operations, existing couriers would be very well placed e.g. for on-site robot maintenance or the remote surveillance and operation of the robots. In sales, existing couriers could leverage their knowledge of demand patterns and other consumption trends in e.g. partner restaurant acquisition. When it comes to roles related to product development, e.g. design and improvement of the service process, user experience and the human-robot interaction, these might also be great options for existing couriers to apply their skills. Finally, moving into more technical roles in both electrical and mechanical engineering or software development and data analytics might require some further training but could form an exciting career development opportunity for existing couriers.

As automation of certain kinds of tasks changes the nature of work (Tuomi et al., 2020), it is important not to ignore the implications technology adoption has on individuals most impacted by it, i.e. those technology might replace. Recognising the pivotal role of existing workers in co-designing new products and services but also in reformulating job roles and tasks as well as highlighting potential training needs becomes extremely important as automation in on-demand delivery marches forward. Indeed, we advocate that platform participants, e.g. couriers, are a pivotal stakeholder of the platform ecosystem, and truly customer-centric offerings can only be developed with platform participants’ (e.g. couriers’) operant resources – skills and knowledge, e.g. spatial navigation. Couriers know what real customers care about and therefore data-centric companies which rely on big data must find ways to incorporate small data (ethnographic data) into their algorithms.

Prior research has noted a similar process when service robotics have been introduced to other service sector tasks: in hotels, check-in robots enable receptionists to move from operations to supervising the robot and only dealing with fringe cases, and in restaurants, food runner robots enable servers to focus more on differentiating the brand by showcasing their hospitality and product expertise, innovating the service process, as well as solving any problems that might arise (Tuomi, Tussyadiah & Stienmetz, 2020). Platform economy companies operating in on-demand delivery should take note and start proactively thinking how to leverage their existing couriers’ domain expertise once the robots hit the sideways in Finland.

In the coming years, the on-demand food delivery services market size and competition are poised to continue to grow, and due to the high couriers’ turnover rates, which characterise the platform economy companies, it is not unthinkable to see shortage of couriers. For example, leaked Amazon memo warns the company is running out of people to hire. In order to provide more streamlined services while respecting fair labour practices and community norms and obligations, platform economy companies need to keep humans in-the-loop in order to design ethical algorithms.


Tuomi, A. & Ascenção, M.P. 2021. Intelligent automation in hospitality: Exploring the relative automatability of frontline food service tasks. Journal of Hospitality and Tourism Insights. ahead-of-print.

Tuomi, A., Tussyadiah, I.P., & Stienmetz, J. 2020a. Applications and implications of service robots in hospitality. Cornell Hospitality Quarterly 62, 2, 232-247.

Tuomi, A., Tussyadiah, I., Ling, E., Miller, G., & Lee, G. 2020b. x=(tourism_work) y=(sdg8) while y=true: automate(x). Annals of Tourism Research, 84, 102978.

More information:

Haaga-Helia’s AlgoAmmatti – Algorithmic Management and Professional Growth in Platform Economy -project seeks to understand algorithmic management practices and the impact of these on workers’ day-to-day experience in the context of digital labour platforms, e.g. Yango, Wolt, or Skillshare. The aim of the service design project is to develop a worker-centric model for conceptualising algorithmic management in the context of professional growth. We seek to create new value for service companies by shedding light on the broader impacts of algorithmic management on digital labour platforms and thus, help companies to proactively develop their services. From a worker-perspective, the goal is to facilitate and manage service work in a more human-centric and socially sustainable manner, focusing on creating balanced and fulfilling careers.

The project is funded by the Finnish Work Environment Fund between 03/2022-12/2023 and conducted by Haaga-Helia’s Service Experience Laboratory LAB8.