In most scenarios, to fully understand how organizations work, it is fundamental to study the interplay between the collaboration among process participants and the environment in which they operate. To this aim, we propose BEAR 2.0, a tool for modeling, animating, and debugging environment-aware BPMN collaborations. This version of the tool enhances its predecessor mainly for what concerns (i) the supported environment model, extended with a logical layer to provide abstraction and diferent levels of granularity, and (ii) the integration of real-world geographic maps, simplifying both the modeling of the environment and the understanding of the process.
In modern organizations, the environment in which processes take place plays a central role in shaping
how participants behave, interact, and make decisions, e.g., a room’s temperature may determine the
next activity to perform in a home automation system. In turn, a participant may afect the environment
and change its status, e.g., by closing a window. Nevertheless, existing modeling languages, including the
de facto standard Business Process Model and Notation (BPMN) [
In this regard, in [
In this paper, we introduce BEAR 2.0 (referred to simply as BEAR from now on), an enhanced version of the tool that provides a modeling environment for designing environment-aware BPMN collaborations. This version of the tool improves its predecessor mainly in terms of the supported environment model, which integrates a logical layer to provide abstraction and diferent levels of granularity, and the integration of a geographic map, therefore simplifying the modeling, animation
(a) Environment-aware BPMN modeler.
Figure 1: The BEAR Modeler GUI. and debugging of environment-aware BPMN collaborations. These advancements allow users to design process models that are grounded in real-world environments.
It is indeed well known from the literature that animator tools play an important role in easing the
understanding of the behavior of business processes [
The rest of the paper is organized as follows. Section 2 describes BEAR functionalities. Section 3 assesses its maturity via case studies. Section 4 concludes the paper by providing links and information about downloading, installing, and using BEAR.
(b) Environment (map).
(c) Environment (graph).
This section presents BEAR and its features for supporting the understanding of the interplay between collaboration business processes and the environment. BEAR provides modeling, animation and debugging functionalities. Modeling features allow the designer to create environment-aware BPMN collaboration diagrams, while animation and debugging are used to visualize and inspect the collaboration execution and the evolution of its environment.
Modeling. The designer can use the modeler to create and link the BPMN collaboration diagram
and the environment model, as shown in Figure 1a. By clicking on any element in the two models, a
property panel opens, which permits adding information related to the selected element and using
them to drive the animation. Concerning the environment, BEAR supports a representation of both the
physical and logical layer in which collaborations take place. The physical layer includes individual
locations where participants operate, called places, and their connections, called edges. Each physical
place and edge can be enriched with a set of key-value attributes, e.g., available = true. These
attributes can be read and modified by the collaboration model using the element name as a prefix, e.g.,
place1.available. This enables processes to both react to and influence the environment during
execution. Attributes can be either spatial, referring to tangible dimensions, e.g., the length of a corridor,
or contextual, describing semantic properties, e.g., the purpose of a room. Spatial attributes enable the
definition of the area covered by each place, allowing for diferent levels of granularity in the model. As
a result, a place can flexibly represent anything from a single indoor room to an entire outdoor parking
area. The physical layer can be directly drawn on a geographic map, but the tool also allows visualizing
its underlying place graph, as shown in Figures 1b and 1c, respectively. This dual representation plays a
key role in supporting the designer’s understanding of the environment, as the place graph highlights
the topology of the physical layer. Building on this, the environment model includes a second layer:
the logical one. This consists of logical places and views, which abstract over the physical layer by
grouping places dynamically based on their attributes. Logical places are defined by boolean expressions
that predicate over environmental attributes. For instance, in the context of a hospital, the expression
available == true && department == radiology may define a logical place called Available
Radiology Rooms. Since attributes may change over time, logical places are inherently dynamic. Logical
places can be directly referenced in the BPMN collaboration model, thus streamlining the modeling
of environment-aware BPMN collaborations. Views further enhance the model’s expressiveness by
enabling interaction at diferent levels of granularity. A view is defined by specifying a name and a set
of logical places it includes. For each attribute of the included places, an aggregation function can be
assigned to compute a collective value. For instance, the attribute availableBeds in multiple rooms
can be aggregated using the SUM function to compute the total number of available beds in a building.
Concerning the collaboration, the property panel shows for each BPMN element, beyond the already
existing properties like element name and identifier, additional information related to the interplay
with the environment. In particular, selecting a pool element, a position attribute can be set, stating
the initial place of the participant represented by the pool. This property appears in BEAR as a text box
to be filled with a place id. About BPMN tasks, the designer can specify (i) guard conditions, used to
constrain the execution of a task to specific values of environmental attributes, which has to be defined
with an expression, e.g., place1.available == true, (ii) a list of assignments, used to modify
environmental attributes values, possibly afecting logical places. In addition, the designer can specify
three diferent types of tasks. (i) Movement tasks, characterized by an arrow icon, which can define
guards, assignments and also a destination, used to define a place to reach, e.g., Room 123, from the
current participant’s position. (ii) Binding tasks, graphically represented with a handshake icon, which
can be linked to a binding task in the pool of another participant, enabling one to follow the movements
of the other. (iii) Unbinding tasks, graphically represented with a two-separated hand icon, which breaks
the binding between two participants, enabling them to resume independent movements. For what
concerns events, the designer can define intermediate message events with value passing. Specifically,
send events can exhibit a payload in the form of values, e.g., “Payload” or 3; or attribute names,
e.g., place1.attr. Instead, receive events can exhibit an attribute name where to store the received
payload. Finally, boolean expressions can be assigned to the outgoing sequence flows of XOR gateways
to drive decisions dynamically, which can also depend on the value of environmental attributes.
Animation. The interface of the BEAR animator, shown in Figure 2, is divided into three main parts,
i.e., the BPMN diagram on the left (highlighted in green), the environment model in the center
(highlighted in red), and the data panel on the right (highlighted in light blue) displaying elements
from both the physical layer and the logical layer, with their associated environmental attributes. The
key characteristic of the animator is its dynamic visual animation. It shows, in a step-by-step fashion,
the evolution of the BPMN collaboration with its environment. By selecting the Token Simulation toggle
in the interface, the environment model will be frozen in its current state and a play button will appear
over each fireable start event. When a button is clicked, the respective process is activated. This creates
a new token, represented as a small colored circle, at the start event of the BPMN process. Another
token of the same color is placed in the corresponding location of the environment model, based on the
position property set on the pool. Then, the tokens start to cross the BPMN collaboration model and
the environment model according to the operational rules induced by our formal semantics [
support efective debugging of environment-aware BPMN collaborations by enabling the designer to identify undesired executions.
Warnings, depicted in Figure 3 in yellow, indicate potential deadlock situations, i.e., two participants attempting a binding task from diferent positions, a participant unable to reach a specific destination, or a violated guard. Animation continues, as these conditions may eventually resolve. Instead, errors, depicted in Figure 3 in red, occur when the collaboration cannot proceed, i.e., a participant’s initial position is missing or two bound participants are moving in opposite directions, and cause the animation to stop. If no errors Figure 3: Debugging features. or deadlocks occur, the animation ends when no tokens can move forward. Nevertheless, it can be paused at any time, allowing users to inspect the token distribution across the two models and assess the environment’s state through the right-side panel.
BEAR is a web application written in JavaScript, which makes it accessible from any browser without a
server backend, and hence, runnable in any operative system. It extends the bpmn-js1 toolkit and its
token simulation plug-in [
To assess the maturity of BEAR, we rely on a set of 21 environment-aware BPMN collaborations.
The 10 case studies originally introduced in BEAR 1.0 have been successfully adapted to the current
version of the tool. Moreover, we have added 11 new case studies spanning a variety of domains, e.g.,
agriculture, transportation, restaurants, universities, healthcare, product sales, and logistics, three of
which are inspired by examples from the literature [
The evaluation of these case studies has shown how environmental conditions can change the outcome of the collaboration process. For the sake of presentation, we briefly discuss one of the case studies concerning an Emergency Response collaboration within a hospital and its surroundings. The collaboration concerns four participants: Injured Patient, Ambulance, Emergency Nurse, and Emergency Doctor. It starts when the Injured Patient is involved in a car accident and calls an ambulance, providing its 1https://bpmn.io/toolkit/bpmn-js/ 2https://openlayers.org position. Upon receiving the call, the Ambulance moves to the patient’s position. During this operation, the ambulance’s travel can be seen in both models: its process token waits near the movement task while its environment token moves across the map. However, if the accident occurs in an unreachable location, such as due to a closed road, the tool displays a warning message. Similarly, failing to specify the ambulance’s initial position displays an error message and halts the animation. The collaboration continues with the ambulance picking up and transporting the patient to the hospital. From a process perspective, the tokens wait for each other at their respective binding tasks. When bound, the patient’s token waits at the next unbinding task, while the ambulance executes the movement task to return to the hospital. About the environment, both tokens move across the map, with the ambulance dictating the patient’s movements. Once at the hospital, the ambulance’s token reaches an unbinding task, detaching from the patient, who will be rescued by the Emergency Nurse and the Emergency Doctor participants. By relying on both place graphs and geographic maps, this new version of BEAR simplifies both the modeling of the environment and the understanding of the process.
We provide BEAR under the MIT license in source code form and as an online service. A demonstration video shows the modeling, animation and debugging of the Emergency Response collaboration presented in Section 3. Additionally, a comprehensive user guide is available in the repository, detailing installation and usage, along with a set of environment-aware BPMN collaboration models, each with a short description of the scenario. All resources are accessible at https://pros.unicam.it/environmental-bpmn.