SAPHARI - Safe and Autonomous Physical Human-Aware Robot Interaction - Workpackages
SAPHARI logo

Login Form

IP 287513
call FP7-ICT-2011-7
SAPHARI is funded under the European Community's 7th Framework Programme

Workpackages

The SAPHARI work plan consists of 8 work packages (WP1‐WP8) on research and technological development (RTD), one on dissemination (WP9), and one on management (WP10). 
The goals of the 8 RTD work packages are summarized below.

WP1 Safety

The safety of industrial robots has always been a major concern as they can build up large amounts of static and dynamic energy. Catastrophic consequences can arise in the event of a collision with humans. Traditionally industrial robots have been segregated from humans by placing them in an enclosure. However, if the robot carries out collaborative tasks with humans, this is obviously not a solution. Consequently, recent research in robot safety has been directed towards making robots safe to use in the vicinity of humans. This WP makes the human the central entity of the evaluation of safety in robotics. We argue that if human biomechanics are not taken into consideration, a realistic prediction of injuries resulting from a collision or the benefits of a particular countermeasure are not possible.

WP2 Building a New Generation of Robots for pHRI

In this work package we will pursue research into development of technologies and components for a new generation of robots that can safely co-exist and co-operate with people, and at the same time get much closerto human performance than is possible with current robots. New technologies will be explored for actuators, sensors, and systems integrating the “robot companion” vision central to of this project. We envision that new technologies will be necessary at all levels of the robot companion architecture, from its higher and more abstract level of cognitive functions to the level of physical embodiment of reactive behaviours in the robot’s actuators and sensors.

WP3 Control

This work package deals with the use of multiple (exteroceptive and proprioceptive) sensing information in real-time control strategies that ensure collision avoidance, detection and fast reaction to unanticipated collisions, as well as a safe continuous physical interaction with the human and the robot in intentional contact. In addition, an important part of the work is devoted to the design and analysis of motion/impedance/force control laws for compliant manipulators based on Variable Impedance Actuation (VIA).

WP4 Perception and Sensor-based Task Monitoring

Goal of WP4 is to develop robust monitoring concepts and components which enable the robot to detect as early and complete as possible dangerous situations in a weakly structured dynamic environment. Since the robot has to share its working space with the human it should be able to interact with him and to act independently (autonomously) without interfering with his actions. Therefore, both the presence of uninvolved moving persons or technical objects (e.g., human performing splice positioning task while the robot is drilling as considered in the use case in WP8) as well as humans and work-pieces which physically cooperate with the robot (e.g., human and robot performing a fastening operation as assumed in the use case in WP8) will be considered. 

WP5 Learning and Interpretation of Human Activities

This work package will provide the methods for recognizing and interpreting the activities of a human partner and for transferring human skills to robots, as required for intuitive multimodal interaction and implicit coordination between a robot co-worker or companion and a human user. The tasks in this work package address the problems of gesture/motion/interaction recognition (using stochastic models, e.g., HMM, that capture both the spatial and temporal variability and by finding the most likely primitive), future movement prediction (both concerning the prediction of discrete motion primitives and of continuous trajectories), as well as context-based task reasoning (using a joint probabilistic model, e.g., BLN, and a task model associating objectives, actions, and perception).

WP6 Reactive Action Generation

With the recent advances in physical Human-Robot Interaction it becomes increasingly important to plan and execute complex motions for task achievement in the vicinity of humans. A major open problem that is especially important for pHRI is motion generation for dynamic environments that solves the completeness-versatility polarity. This means finding a feasible path if one exists but at the same time planning fast enough to change it in accordance to dynamic changes of the environment. Up to now there exists no motion planning algorithm capable of dealing with the proximity of dynamic obstacles, possibly under the absolute premise of human safety. In WP6, we will try to overcome the traditional approach of separating motion generation from control and design entirely new solutions.

WP7 Human-aware Plan-based Control for Collaborative Tasks

The objective of WP7 is to endow the robot with the task planning, monitoring and control abilities that are necessary to share its space and task with a human partner, involving also close physical cooperation as a core element. In this context, the robot should not only ensure the feasibility of its part of the joint task but also satisfy safety and legibility requirements of its motions. Therefore, reasoning mechanisms should not only involve task and action management but also include spatial reasoning on the environment, on the humans, on the objects to be manipulated, and on the robot itself. Globally WP7 has to do with so-called higher-level context/task based decision layer for SAPHARI robots.

WP8 Case Studies and Prototype Scenarios

It is the target of SAPHARI to produce results that are highly relevant to all stakeholders. Academia will be interested in the scientific challenges of WPs 1-7. The first objective of WP8 is to integrate most of this work and to evaluate the performance in realistic use cases. The second objective is to ensure that the robotics industry becomes interested in the outputs of the project and that the technologies developed find their way into products. For this reason the industrial requirements will be analysed early and industrially relevant integration platforms will be turned into demonstrators highlighting the most mature technologies developed within the project.
Institute of Robotics and Mechatronics (DLR)