Aspects Influencing Users' Trust in an Automated Delivery ‘Bot’ (ADR): A Pilot Study

Conference poster, 2023

1.     Purpose
Autonomous Delivery Robots (ADR), i.e., vehicles that require minimal to no operator interference and that uses system mapping, GPS, stereo vision, sensors and edge computing to navigate itself from one location to another (Mishra & Das, 2019), have been introduced as a possible solution for last-mile deliveries (LMDs) of parcels and goods. LMDs are the final stage of the logistics chain and considered one of the most inefficient (as well as the most expensive) part of a logistics system (Sindi & Woodman, 2020; Sonneberg et al. 2019). However, before logistic service providers and logistic personnel can reap the benefits of bots for LMD, it is important to understand how this novel technology is received and trusted. Therefore, a pilot study was initiated with the purpose to explore how logistic personnel experienced and trusted a bot as a tool for LMD.
 
2.     Research Approach
The bot was implemented into the logistic service system at a university campus in Gothenburg, Sweden. The research design used included qualitative, semi-structured interviews and questionnaires distributed to participating logistic personnel both ex-ante and ex-post trial. Topics covered included expected vs. experienced trust but also perceived benefits, workload, as well as overall experience of using an ADR for LMD at a university campus.

3.     Findings and Originality
Whereas most studies on trust in automated artefacts often focus only on trust in the automated artefact as such, the pilot study showed that the users’ trust in the delivery bot was not only a matter of trust in the technology or the automated artefact per se. Based on the interview and questionnaire data, the participants' trust in the bot was influenced by a number of interdependent system factors: (i) the perceived performance of the bot, (ii) perceived risks in terms of possible theft and traffic accidents, (iii) the degree to which the bot influenced perceived task difficulty and workload, (iv) the perceived benefits of the bot as part of the logistic system, as well as (v) the perceived benevolence behind the development and introduction the bot, from management as well as the designers of the bot.

4.     Research Impact
Little research has been conducted on user trust in ADR technology and even less focus on accounting for contextual aspects in the overall logistic system and their effect on user trust. Therefore, the presented research shows the importance of taking a holistic approach towards trust in ADR when introducing it to the market.

5.     Practical Impact
Introducing and implementing new technology such as ADR into a service is not a simple “plug and play” process. Instead, stakeholders responsible for the implementation needs to be aware of the specific needs of the user(s) i.e. people that on a daily basis interacts with the ADR such as logistic personnel.
 
To support the introduction and implementation of ADR there needs to be a dialog between management and logistic personnel but also between developers and logistic personnel. A dialogue crucial to mitigate risks for over-under trust, acceptance issues but also to avoid negative consequences such as sub-optimal logistic services.
 
6.     Discussion Questions
1.     What is needed to support the introduction of new technology such as ADR, so ADR becomes something positive for everyone?
2.     How can we support testing of novel technology such as ADR to be able to capture important factors that affects e.g. user trust in automated systems?
3.     Who should have the responsibility for the development of logistic technology that entails advantages not only for the employer/management but also for the logistic personnel?