Minimising risks for seasonal fruit pickers
Through Human-Technology Integration

By Cranfield University

Foraging for wild food has an endemic and enduring history in Europe. It is one of the oldest forest activities in many countries, particularly in Scandinavia, where fertile growing conditions are provided by long hours of sunlight in summer and large expanses of undisturbed wilderness. Thanks to common access and harvesting rights, numerous edible wild plants, berries, and mushrooms are widely available for anyone to harvest. Commercial wild food harvesting is undertaken on a much smaller scale than domestic picking, but there are clear indications that there is room for expansion in international markets.

Challenges of berry-picking in Finland

To meet the growing demand for wild berries, Finnish berry trading companies increasingly rely on a substantial foreign workforce, employing large numbers of seasonal pickers from abroad. Since 2005, a significant contingent of Thai workers has been permitted to harvest wild berries in Finland during a season that aligns with Thailand’s monsoon period. This timing makes the opportunity particularly appealing, as the income earned in Finland can represent several years’ wages in Thailand. However, migrant pickers assume considerable financial risk, as they bear the upfront costs of travel, accommodation, and related expenses without any guarantee of earning enough to offset these costs (Yle News, 2024). To address concerns about worker exploitation and to improve working conditions, Finland introduced new legislation in 2021, mandating that foreign berry pickers must earn a minimum of 30 euros per day, in accordance with Schengen visa regulations (YLE, 2021). This measure aimed to enhance the economic security of migrant workers. More recently, in response to rising concerns around exploitation and trafficking, Thai pickers are required to secure work-based residence permits, which offer further labour rights protections (YLE News, 2024). Nevertheless, the occupation remains demanding. Earnings are highly variable, contingent on the type, quality, and quantity of berries harvested. As a result, pickers face long, physically challenging, and mentally taxing workdays, striving to maximise their harvest and, consequently, their earnings. While potentially lucrative, this line of work is marked by significant physical and financial pressures.

Firstly, the working day for berry pickers is long so that they can make the most of their time in the forest, typically rising early (around 4.30 am) to travel some distance by road to the chosen forest location, where they will then walk and forage throughout the day, often with little or no breaks and not returning to base before 8.00 pm.

Secondly, the physical strenuousness of fruit picking work is high, as pickers need to constantly move, bend, and stretch over rough terrain to reach and gather berries using combing/cutting tools, which are then emptied into large buckets that need to be carried. Lifting and carrying these buckets gets more taxing through the day as weight increases and as pickers inevitably become progressively fatigued. All this exertion is exacerbated by the warm, highly humid conditions of the rainy forests during the summer season.

Thirdly, cognitive demands are a less salient but nonetheless significant challenge in berry picking work. Pickers must accurately identify the most suitable areas to forage and how to reach ripe berries. Identifying the most lucrative locations to forage is a considerable challenge, even for local pickers. Prime crop locations are estimated using personal experience of weather and geographical conditions. Pickers’ cognitive demands are also heightened by the need to maintain situational awareness and avoid personal injury as they negotiate unfamiliar and uneven terrain.

Unmanned Aerial Vehicles

Unmanned aerial vehicles (UAVs) or ‘drones’ are a frequently studied area of robotics, as they offer unique manoeuvrability, hovering, and low-altitude flight capabilities for various applications and tasks (Valavanis, 2017). Up to now, there has been little or no research and development of UAV systems to support wild food foragers, either commercial or recreational. UAVs are ideal for the photogrammetric mapping of crops when flights are unhindered by obstacles. Still, under-canopy flights between trees and foliage in forest mapping scenarios, either with a human pilot or AI navigation, are currently very challenging. In addition to the many technical challenges faced by drone deployment in berry-picking environments, little research has been conducted to explore the collaborative human-centred aspects of drone-supported berry picking and whether this proposition can deliver measurable improvements to pickers’ productivity, health, and well-being.

Human-robot interaction research has provided valuable insights that have guided the technical design of collaborative robotic systems. We know that human trust (that the robot will make appropriate decisions and behave in an expected and understandable manner) is crucial for effective collaboration (Gil et al., 2019), and levels of trust are influenced by specific features such as perceived robot speed, reliability, and safety (Charalambous et al., 2016). Current knowledge about the impacts of UAV characteristics on the responses of human collaborators, operators, and other individuals in the vicinity of UAVs is unknown. To address this need, the field of “human-drone interaction” (HDI) has been proposed as a dedicated area of investigation for understanding, designing, and evaluating drone systems for human users (Tezza and Andujar, 2019).

In addition to trust, human-robot collaboration is also influenced by a range of specific psychological and affective states that the human experiences in a situation or context, which can be highly instrumental to performance and well-being outcomes. In particular, mental workload, situation awareness and (intrinsic) job satisfaction are key factors directly influenced by system characteristics such as task complexity, level of autonomy, speed, and safety (perceived and actual). The interplay of these human factors and UAV system characteristics has yet to be defined and remains a research gap.

Human analysis

Identifying what the intended users of a new technology want or need is a foundational step in effective system design. In the specific context of developing UAVs to optimise berry picker performance and well-being, where there are currently no directly relevant research findings to guide design, the first task for FEROX human analysis must be to identify their requirements. The experiences and expectations of berry pickers are investigated to establish conventional work practices and identify specific limitations and problems, and to explore how they believe a UAV system should be designed to assist them best and identify preferences and needs. We use interviews, eye-tracking, and online psychometric surveys for this. One recent technical report provides the first international standards document on human-centred aspects of robotics (ISO/TR 9241-810:2020), but this provides a general level of guidance that is not specific to UAV characteristics.

Consequently, without dedicated standards or specifications for this context, the next project objective will be to apply the user requirements data to inform the design of FEROX systems and interfaces. Specifically, the information provided by interviews on user experiences and opinions and the eye-tracking data showing real picking work will be distilled and mapped to provide a usable set of guidelines. The final objective for the human analysis work in FEROX will be to measure the impact of the new UAV solutions and assess how well the user-centred design has successfully improved the work and assisted pickers. The performance and well-being data gathered in the user requirements phase will be followed up by the same or comparable techniques so that changes can be assessed qualitatively or measured quantitatively.

The collaborative approach between humans and UAVs offers the opportunity to improve berry location accuracy, increase overall yield and incomes, and provide monitoring and safety benefits. Furthermore, incorporating robotics and technology into berry-picking endeavours can potentially engage younger generations or tourists and widen the demographic of pickers, creating a sustainable future for this tradition.

References

YLE News (2021). Osa thaipoimijoista keräsi 15 vuoden tulot reilussa kahdessa kuukaudessa. 6 Oct 2021.

YLE News (2024). Firms say berries will stay on bush as Finland suspends Thai berry-picking visas. 22/03/2024. https://yle.fi/a/74-20080493.

Valavanis, K. P. (2017, July). Unmanned Aircraft Systems challenges in design for autonomy. In 2017 11th International Workshop on Robot Motion and Control (RoMoCo) (pp. 73-86). IEEE.

Gil, M., Albert, M., Fons, J., & Pelechano, V. (2019). Designing human-in-the-loop autonomous cyber-physical systems. International journal of human-computer studies, 130, 21-39.

Charalambous, G., Fletcher, S., & Webb, P. (2016). The development of a scale to evaluate trust in industrial human-robot collaboration. International Journal of Social Robotics, 8, 193-209.

Tezza, D., & Andujar, M. (2019). The state-of-the-art of human–drone interaction: A survey. IEEE Access, 7, 167438-167454.