1 Introduction

Nowadays, technology innovation, challenged by industry 4.0 pillars, is pulling the manufacturing improvement [13, 15, 16].

One of the most interesting challenge is the robotics improvement and its applications; in particular, the human–robot collaboration (HRC) tasks. HRC is based on two resources, human being and collaborative robot (cobot), that work side by side [17] to perform a common task.

A cobot is a special kind of robot, environmentally aware, which can share the workspace with other cobots or with human operators. Sometimes it is equipped with vision system, sensitive skin and proximity or contact sensors [3]. They are very suitable for light and repetitive activities in limited space where human presence is allowed. On another hand, they must work with low speed and payload and consequently with a limited productive capability. However, the combination of human and robot strengths allows to achieve more flexible production processes maintaining the repeatability and accuracy of classic industrial automation [2].

The implementation of collaborative robotics is still a challenge due to a significant increase in complexity [8]. Indeed, the human presence in the workplace involves consideration of aspects that go beyond the robotics alone.

In this context, the facility layout problem of the HRC workplaces is one of the most interesting technical challenge [10, 12] to face in order to achieve an efficient collaboration between humans and cobots [4]. Pini et al. [11] propose a systematic approach to identify the resources more suitable to perform the different tasks, but the organization of the layout is left to the designer perception. Other researchers [9, 14] face the task allocation to humans and robots, but the layout design is not deeply tackled.

Furthermore, the above-mentioned approaches do not face the regulatory constraints and the compliance with standards in the methods and tools they have developed. Therefore, this work, through the analysis and the critical study of the reference frame, proposes a knowledge-based approach aimed to help the designer to move among the complex regulatory framework in order to develop a compliance HRC layout.

2 Standards and constraints

Ensuring safety of the human being in workplaces is the main challenge to tackle. The International Organization for Standardization faced the safety problem providing a huge amount of rules and prescriptions strongly interconnected [5,6,7]. However, it is very hard to move among the maze of standards that sometimes are merely guideline [1]. Even though, in the international standards, all the spaces within the workplace are defined, the operating space, that is the space where the task sequence is carried out, is exclusively dedicated to the robot, whereas the collaborative space, where human presence is permitted and the interaction between human and robot can occur, is only a portion of it. On the other hand, the definition of these spaces does not consider the human presence, indeed the collaborative operations described in the standards are all robot oriented, due to safety reasons.

Many other prescriptions and constrains shall be considered from standards and production needs. All these elements can be summarized in the following list: (1) type of operation; (2) physical limits and clearance; (3) environmental conditions; (iv) ergonomic limits; (5) human intervention and tasks; (6) robot parameters; (7) material feature and conditions; (8) workspaces division; (9) safeguarding perimeter and accesses; and, (10) devices and resources.

To comply all these aspects is not easy, therefore the designing is left to the experience of the designer and the commercial software have no constrain implemented to comply the standard prescriptions.

3 Tools

Several instruments and tools can be used to face the challenges about human–robot collaboration workplace design. Through several engineering design methods, such as axiomatic design and graph theory, it is possible to extrapolate the basic concepts of a collaborative workplace, to find the requirements defined by the standards and to link each of them to a design parameter. Then, all this information needs to be organized and managed. The relationships among the elements can be organized by a matrix (graph) to exploit and manipulate this knowledge.

The easily implementation and the large number of implemented codes give this graph-based tool back very powerful. Commercial software, such as Simulink (a MATLAB-based graphical programming environment), provide several algorithms in order to obtain desired results according to the inputs provided.

Finally, it is possible to generate a digital mock-up of the collaborative workplace and explore it in virtual environment in order to perform a design review.

4 The proposed approach and results

This work proposes the layout design and optimization of human–robot collaborative workplaces based on the knowledge of the functional requirements and constraints imposed by the reference standards. The interdependence among the huge amount of design parameters and functional requirements are managed by means a multi-level graph-based approach. The graph-based approach allows easy implementation of the knowledge in a calculation code aiming to obtain the layout of human–robot collaborative workplaces in a semi-automatic way and in standards compliance.

The calculation code, named Smart Positioner, has been developed in Matlab environment. Figure 1 depicts the flow chart showing the functioning of the Smart Positioner. It is characterized by four main steps: (S1) manually inputs insertion—the designer inserts as inputs: (1) cell dimensions, (2) resources characteristics (number, type and dimensions), (3) tasks sequence. (S2) Generation of the HRC workplace layout in a three-dimensional environment. The tool performs a first check on the generated solution to assure compliance with the regulations. The layout visualization is also available in a 2D environment in order to allow easier control of the overall dimensions and visual obstructions. (S3) Customization by the designer: the tool allows the user to evaluate and edit the layout. Then a check occurs to verify the standards compliance. (S4) Conversion of the file to JT (Jupiter Tessellation) format in order to export it to simulation environments. This tool provides a crucial support to the designer to move through the complexity of the huge amount of constraints and requirements of the standards.

Fig. 1
figure 1

Smart Positioner flow chart

Full size image

The resources of the workplace are organized using optimisation criteria due to minimize the floor space occupied, the distance covered by the operator and the material handling costs, under constraint of minimum distances to comply.

Definitely, the proposed Smart Positioner represents a useful tool to reach an efficient human–robot collaboration, facing the layout design problem of the collaborative workplaces by means a knowledge-based approach.