[Context and Motivation] The requirements elicitation process for socio-technical systems requires the involvement of diverse stakeholders with diferent backgrounds and skills. In these contexts, effective communication between business analysts and stakeholders can be supported by model-based requirements engineering (MoDRE) strategies, which leverage diagrammatic notations as a means for information exchange. [Question/Problem] Several diagrams and approaches exist to facilitate MoDRE. However, empirical evidence on their applicability to real-world contexts with a relevant social component, and going through a process of digitalisation, is limited. Furthermore, existing approaches do not evaluate the impact that the deployment of a novel digital system has on the process and its actors. [Principal idea/Results] The research outlined in this paper aims to evaluate the joint usage of typical requirements engineer notations, namely i*, class diagrams, and business process models in the elicitation of requirements for socially-intensive systems that are going through a transformative digitalisation process. We apply these notations to represent the system as-is, and the system to-be, with the goal of also evaluating the impact of digitalisation. We focus on living labs (networks of stakeholders in a socio-technical system) belonging to the agriculture domain, and provide a preliminary application on a farm that is introducing an AI-based irrigation system. [Contribution] The results show that efective communication with non-technical stakeholders is feasible with the envisioned approach. Although multiple iterations are required, agronomists and farmers are able to provide constructive feedback on the basis of the models. Furthermore, impacts in terms of additional/removed tasks and actors can be efectively characterised through business process models. As part of our overall project, we will refine the method, and then apply it in 20 living labs in the EU.
Digitalisation in agriculture is a socio-technical process to be considered from multiple
perspectives, such as social, economic, environmental, and technological. The focus on the introduction
and adoption of Information and Communication Technologies (ICT) — a specific class of
technical tools — has brought to the development of the concept of socio-cyber-physical systems
in [
The reference conceptual framework is Ostrom’s socio-ecological System (SES) framework
[
A key activity will be process modelling, aiming at the creation of easy-to-read graphical
representations of processes in FASs to describe and evaluate how the introduction of digital
technologies may reshape a context (i.e., their impacts). The results of the analysis will be used
as input for further activities, such as the description of DEs and the cost-benefits analysis.
The models will represent a fraction of the FAS each LL will define, and they will be focused
on the re-engineering of the processes, aiming at the representations of the processes before
and after the introduction of digital technologies. The process modelling will be developed by
applying MoDRE strategies. This will contribute to filling the lack of empirical evidence on the
applicability of MoDRE to real-world contexts with a relevant social component, as evidenced
in [
The overall objective of the project is to perform an analysis of DEs from multiple
perspectives. However, both the transdisciplinarity of the research and the direct involvement of
LLs challenge the communication at diferent layers: among diferent academic disciplines
(economics, sociology, ecology, engineering, etc.); among diferent actors (researchers, farmers,
intermediaries, stakeholders, policymakers, etc.); among domains and specialisations (water
management, nutrient management, crop protection, livestock, etc.). Therefore, the purpose of
our study is to address these research questions:
• RQ1: How to perform requirements elicitation in socio-technical systems in order to
create representations that account for social relations and goals, structural elements,
processes, and impacts of digitalisation?
• RQ2: What is the degree of user acceptance (usefulness, ease of use, and intention to use)
of the devised elicitation procedures and system representations?
To answer RQ1, we will: 1) select graphical models from the MoDRE field, considering prominent
notations for the representation of social interactions and goals, system components, processes,
and impacts, and adapt the notations to maximise usefulness and understandability; 2) develop
structured communication procedures between requirements analysts and stakeholders, with
the goal of capturing elements of the systems and relations thereof, as well as impacts due to
digitalisation. The objective of these procedures is to enable the representation of the system
and its transformation according to the selected MoDRE notations. To achieve these goals,
we will follow the design science paradigm [
The activity is based on the application of MoDRE strategies to understand how current
processes are re-engineered after the introduction of digital technologies. The transformation is
emphasized by qualitatively highlighting the diferences in the process as-is (before) and in the
process to-be (after). To ensure completeness, a set of languages has been so far identified and
the following diagrams will be developed by each LL following a guided procedure:
• the UML class diagram will provide an overview of the classes, i. e. actors, tools and
infrastructures involved in the process to-be, and the relationships among them [
The i* and UML diagrams focus on the process to-be only to simplify the overall representation, considering that the core models to visualise the process transformation are the BPMN diagrams. This was a pragmatic decision of the authors.
In order to assess the modelling methodology, a pilot study based on a smart irrigation system adopted on a pear orchard by Illuminati Frutta2, a fruit farm in Tuscany, was analysed and diagrams were drawn. The adoption of a precise irrigation system, composed of a wireless sensor network (WSN) and a decision support system (DSS), has promising potential in terms of economic, productive and environmental benefits. Figure 1 contains the UML class diagram which provides an overview of the classes, i. e. actors, tools and infrastructures involved in the precision irrigation system and the relationships between them, i.e., roles and actions performed. The new classes introduced by digital technology are in light blue. In Figure 2, the i* diagram represents the goals of the irrigation process. Highlighting actors, components and activities in relation to the goals and qualities, the model focuses on the demonstration of the positive impact of the introduction of the new actor “DSS infrastructure” and its measurement activities in supporting the irrigation task of the farmer. Figure 3 is an overlap of two BPMN models: a ifrst diagram describing the process before the introduction of technology, and a second diagram representing the precision irrigation system introduced by the DSS infrastructure. In the picture, the new participants, activities and gateways introduced by technology are in blue; in green the items that do not change; in red the items that are not present anymore (an accessible version of the picture, along with an animated version and the single process diagrams, is available in the replication package— see Footnote 1). The aim of this visualisation is to provide an overview of the process re-engineering impacts. This is expected to be used for an in-depth qualitative discussion on the efects of digitalisation (novel/removed tasks, actors, consequences of the transformation, etc.) with the involved stakeholders.
The procedure adopted to define the models is as follows: • The authors of the paper (computer scientists) visited the farm and dedicated half a day to interacting with a group of ten people who consisted of farmers, agronomists and the owner of the farm. 2http://www.illuminatifrutta.it last visited 9 February 2023 • The agronomists were asked to write a document in natural language describing the system structure and the automated process currently under development, especially describing the change with respect to a manual process. • Based on the document, the first author created the diagrams. These were revised with experts in i*, class diagrams and BPMN, and this led to 11 changes to better comply with the grammar of the notations (all changes are tracked in an Excel file, available in our replication package—see Footnote 1). This suggests that, also in future instances of the methodology, this step requires the involvement of experts to ensure that the diagrams are syntactically correct. • The diagrams were used to further clarify certain aspects of the system in a 2-hours meeting with the agronomists. This led to 2 substantial changes in the diagrams, (new components and relationships in the UML), and the meeting showed that the agronomists clearly understood the notations and were able to provide feedback. The agronomist particularly appreciated the BPMN diagrams, but highlighted that the representations did not consider the process for setting the setup of the new system, not allowing for an analysis of the costs of the transition (e.g., training, pilot tests). Despite that, it was agreed not to include additional diagrams, in favour of leaner models focusing on the representation of the system before and after the introduction of digital technologies. The representation of the transition will however be considered in other tasks of the project.
The current stage of our research consists of the requirements definition step of the design science treatment targeted to RQ1. The pilot study is the output of the outline artefact sub-activity aimed at identifying the type of artefact to develop; it includes the diagrams presented above and a proposal of process modelling methodology. Following the project timeline, the feasibility study was carried out as a preliminary activity and was evaluated during a focus group with project partners and a plenary meeting with LL coordinators. The evaluation allowed to complete the sub-activity of requirements elicitation and the positive feedback reafirmed the potential of MoDRE to produce standard and understandable representations of processes to be used in further analysis by stakeholders with diferent backgrounds. The methodology to carry out process modelling is based on the same procedure adopted for the development of the models. Some adjustments were made to comply with two major project constraints: 1. data collection will be performed in coordination with other tasks 2. limited number of available iterations necessary to refine the models. On the one hand, iterations are fundamental for precision and high levels of detail in the models; on the other hand, they are time-consuming and require the availability of multiple actors. Eventually, a five-step methodology to carry out the process modelling has been agreed upon. The methodology is based on these steps: 1) data collection; 2) reporting; 3) check; 4) formalisation; 5) agreement. Steps 1 and 2 will be performed by LLs together with LLs coordinators, steps 4 and 5 will be carried out by task leaders with LLs’ support, while step 5 will be the final agreement between task leaders and LLs. The current task is the production of guidelines for LL coordinators in charge of data collection. Clear guidelines describing the activities to be carried out are fundamental to collecting information for applying RE techniques in the representations of the systems. Thus, the next stages will consist of the design and development of the artefact, and the demonstration. As a design science project, it will be carried out in an iterative manner, moving back and forth between all the activities, adapting the models and the procedures to LLs needs and refining the overall process. The final phase will be devoted to answering RQ2, with the evaluation both of the system representations, and the overall methodology through standard questionnaires administered to the LLs.
This paper presents preliminary results arising from the ongoing research project EU CODECS. LLs involvement helps both to address emerging challenges related to the adoption of ICT and to perform research in real agricultural contexts. Diagrammatic notations are the common language for information exchange between diferent actors, and MoDRE strategies help to describe socio-technical systems. On the one hand, having a clear representation of the processes tailored to farmers’ needs is fundamental to performing the technology assessment of such systems; on the other, the focus on the process re-engineering will allow describing the impacts of the introduction of digital technologies. Future challenges include the development and evaluation of a successful methodology for data collection and modelling, and additional models to tackle the transition phase between the before and after the introduction of technology. This work has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement no. 101060179.