5. Assisted Navigation

Human-Machine Collaborative Control

Objectives:

• R&D of a new human-machine collaborative control strategy for the navigation of assistive mobile robots commanded through the use of multimodal HMIs, that can either be continuous or discrete, and taking into account situation awareness, semantic, and user´s state of mind.
• R&D of a new planning strategy to allow robust navigation of non-circular differential mobile robots in dynamic changing human environments.
• Design of a complete assistive navigation framework for ROS environment.

Future Work:

1. To extend the current collaborative control approach to allow:
   • Integration of human user state of mind in terms of stress, fatigue, focus, and emotion;
   • To improve the situation awareness module to allow online settling of local navigation goals (e.g. to dynamically identify specific goals in a room such as a table).
   • To allow robust navigation of the assistive robot through the use of multimodal interfaces that can either provide discrete (synchronous or asynchronous) or continuous command signals.
1. To continue the research and development of a novel path planning approach for large non-circular differential mobile robots operating in indoor unstructured environments.

Video Experiments with RobChair:

Dataset Experiments with RobChair:

Data Analysis (source code)

Publications:

• [IJ16] A.C. Lopes, G. Pires, J. Perdigão, J.P. Rodrigues, U.J. Nunes, A New Hybrid Motion Planner applied in a Brain-actuated Robotic Wheelchair, IEEE Robotics and Automation Magazine, Vol. 23 (4), Dec. 2016.
  • DOI: 10.1109/MRA.2016.2605403
• [MSc5] Diogo Gonçalves. RobChair 2.0: Simultaneous Localization and Mapping and Hardware/Software Frameworks. Master’s Thesis. Institute of Systems and Robotics, University of Coimbra, September 2013. (supervisors: U. Nunes and A. Lopes)
  • [MSc6] Jorge Perdigão. Collaborative-control based navigation of mobile human-centered robots. Master's thesis, Institute of Systems and Robotics, Institute of Systems and Robotics, University of Coimbra. July, 2014. (supervisors: U. Nunes and A. Lopes)
  • [TR3] João Rodrigues and Jorge Perdigão. CollabNav01. Technical Report. Institute of Systems and Robotics, University of Coimbra. May, 2015. (supervisors: U. Nunes and A. Lopes)

Active Safety and Assessment of User Needs

Objectives:

• R&D of an active safety system based on 2.5D and 3D depth data provided through a Kinect sensor.
• Characterization of mobility, accessibility and safety of severe motor impaired users and assessment of these user needs, especially those requiring a powered wheelchair for mobility.

Main achievements:

• Development of a low cost active safety system based on 2.5D depth data provided through a Kinect sensor [MSc7, MSc9]
• The research work presented in [IC10] characterizes mobility, accessibility and safety of individuals with severe motor impairment such as users suffering from Cerebral Palsy (CP). The sample was collected in Coimbra Cerebral Palsy Association (APCC) and it included 16 individuals with CP. To these individuals we gave an evaluation protocol with a form with clinical and socio-demographic data and a questionnaire on mobility, accessibility and safety, especially concerned with the use of powered wheelchairs. The main limiting factors expressed by these individuals include building/vehicle access, difficulty in reverse drive and lack of safety. The most valued features of a powered wheelchair are comfort and structure, easy navigation and wheelchair control and safety. The lack of safety in the outdoors was also a relevant limiting factor. Almost all individuals requested improvements of the powered wheelchair. The most requested improvements were safety related or related with navigation problems. An assistive navigation solution based on a shared control algorithm is presented, where a powered wheelchair is equipped with the Kinect sensor, in order to help the user maneuvering the wheelchair safely [MSc7].

Video:

Experiments Active Safety: Steering a powered wheelchair with a joystick

Future Work:

• Improve the robustness of the planning approach used on the active safety system based on 2.5D depth data developed in [MSc7, MSc9].
• Assessment of the proposed solution in real indoor scenarios by able-bodied and motor disabled users.

Publications:

• [IC10] Ana Marta Carvalho, Alireza Asvadi, Carlos Carona, Ana Lopes, Urbano Nunes, Mobility, Accessibility and Safety of People with Cerebral Palsy, HEALTHINF (BIOSTEC 2015), Lisbon, Portugal, January, 2015
• [MSc7] Ana Marta Carvalho. Reactive Navigation based on 2.5D Kinect Data for Safety Improvement in Mobile Assistive Robot Steering Master’s Thesis. University of Coimbra, September 2014. (supervisors: U. Nunes and A. Lopes)
• [MSc9] Hugo Paiva, Human machine interface modules to support indoor navigation of a robotic wheelchair. University of Coimbra, September 2015. (supervisors: U. Nunes and A. Lopes)

5.3 Assessment of the Assistive Navigation System in Real Indoor Scenarios

Objectives:

Main Achievements:

• Four navigation tasks were established to test the effectiveness of the ANS in a real office-kind environment.
• The following figure shows the navigation task scenario and the navigation tasks designed to assess the system robustness and efficiency.
• The experimental navigation tests were carried out by ten disabled participants and by ten able-bodied participants [TR2, sIJx1].
• The the self paced BCI, as well as the non-self paced BCI were used as system interfaces.

Assessed navigation tasks with RobChair setup using a self-paced BCI in a real office-like scenario. Four navigation tasks are considered TASK 1 to TASK 4. WC and EXIT are global goals, and START, LAB B and OFFICE C are local goals. Points A and C to M regard decision points associated with intersections. Points O, P and Q denote glass windows, and point N is in respect to a user change of plan.

Future Work:

• Definition and application of benchmarks and metrics to evaluate the data obtained during the course of the described experiments with RobChair.