Levels of interaction

by Sep 13, 2021Advanced manufactoring solutions0 comments

Aside from the physical interaction, which by definition cancels the distances and separations between operators and robots, the technologies for sharing the workspace will be aimed at transforming traditional companies into a more fluid and dynamic, open and interactive environment (measurement and monitoring of the environment and operators, dynamic allocation of programs, navigation of mobile platforms, secure virtual spaces, multi-robot technologies).

At the base of the interaction technologies, you will find all the technologies and aspects relating to the safety of physical interaction and sharing of spaces. Specifically, cooperation can be characterised differently (or can be defined) at various levels such as:


(physical Human-Robot Interaction, pHRI): where there is a direct exchange of energy between the human operators and robotic agents, for example in joint manipulation and contact (intentional or accidental). Examples of such a method include:

  • intuitive programming (lead-through programming) in which a manipulator is trained by accompanying the movement, physically guiding the manipulator along trajectories, to be then repeated independently later;
  • the simultaneous handling of the same parts (material handling) for large loads or particular positioning;
  • the possibility of limiting/stopping the motion or the execution of automatic tasks by directly blocking the manipulator intentionally or for emergency interventions;


where the organisation of the production space provides for competition between both human and robotic operators. The collaboration can take either serial modes (production workflow steps alternating between robot/operator) or parallel (independent operations joined to certain steps). In the case of shared tasks, these must be carried out in collaboration with the sole purpose of achieving maximum effectiveness or in cases where simply replacing the operator does not give any added value or the process cannot be achieved otherwise. Examples include specific assembly cases, in which the robot co-manipulates large parts or prepares guides/assembly aids. In the case of parallel processes, the organisation of the shared workspace (workspace sharing) makes use of motion rescheduling technologies (collision avoidance, speed reduction, target reallocation) and of the task (adaptive rescheduling) in order to guarantee the shared space safe for the operator;


complementary to the previous ones, where the organisation of shared processes provides for a certain degree of understanding of the context (context awareness). The transparency of interaction between man and robot is all the greater
the greater the extent of the transfer of activities, commands, explicit information from the operator to the robot. That is, the more the robotic agent begins to interpret and manage information locally (e.g. specific tasks in relation to a specific processing step, program parts in relation to a specific cell configuration, specific configurations in relation to particular contexts of risk), the less the burden of explicit redundant transfer of notions from the operator to the machine with consequent reduction of execution times, improvement of the user experience, greater flexibility and speed of reorganisation of tasks. This component acts not only at the device level (robot) through a series of behaviour and gesture recognition technologies (typically including three-dimensional vision sensors, reconstruction of environments and of the human counterpart, model-based matching, machine learning, deep learning, etc.). It also works at the system level where the single information generated locally is distributed and shared with other involved agents (process planning, system supervision, resource allocation, etc). To make this organisation of information possible, technologies of an architectural level are needed, including both infrastructures (control system architectures, knowledge-base control) and modelling components (environment modelling, behaviour modelling, sensor data modelling, object modelling).

The various levels of cooperation will contribute to building a system of sensory, motor and cognitive-functional skills such as to create, in the 4.0 Factory context, a community of mixed agents (human and robotic) whose flexibility in the allocation of activities, behaviours and tasks will be very high.