Delft Haptics Lab

Delft University of Technology

VIDI: Symbiotic Driving

Principal Investigator: David Abbink
2016-2021 (VIDI Research Programme, Active)
Abstract

The research in this programme is based on a personal grant proposal (VIDI), which was awarded to David Abbink in May 2015 by Dutch funding agency TTW, part of the Dutch Science Foundation NWO. It provides funding for David, two PhD candidates and one Postdoc, who will combine their research efforts together with DAVI, Moog and Nissan, in order to develop “a feel for symbiotic driving”.

Symbiotic Driving is an alternative to the traditional “traded control” approach (driver-as-a-backup-system) of the automotive industry for driving intelligent vehicles. With symbiotic driving, we propose a mutually adaptive and learning interface based on haptics, where driver and vehicle co-operate, support each other and learn from each other through physical interaction.

The project officially started in spring 2016, with the hiring of the first PhD student.

Project Members

  • Project Leader: David Abbink
  • PhD on cybernetic driver modeling and identification, for real-time adaptation of haptic shared control: Sarah Barendswaard
  • PhD on modeling the underlying principles that govern driver adaptation Sarvesh Kolekar
  • Postdoc on design and evaluating adaptive shared control prototypes in simulators and on test-tracks – Bastiaan Petermeijer
  • User Committee: Nissan, DAVI, Moog

Proposal Title: Developing a feel for symbiotic driving – Establishing haptic shared control as adaptive, individualized interaction between driver and highly-automated car

Proposal Abstract: The question is not if automated vehicles will be introduced on our roads, but when and how. A major bottleneck is communication and interaction between driver and highly-automated vehicle, especially when drivers need to regain control quickly and unexpectedly. Conventional human-automation interaction (beeps, switches) is inadequate, but academia and industry are struggling to invent better alternatives.

A promising approach is haptic shared control, where drivers can feel (and interact with) the automation’s control activity through forces on the steering wheel or gas pedal, similar to the intuitive physical interaction between horse and rider through the reins. However, where horse and rider establish a symbiotic relationship, today’s haptic interfaces fail to adapt to highly individual and situated driver-behaviour-preferences and automation reliability. Their ‘one-size-fits-all’ design causes conflicts between driver and haptic automation, which in turn reduces driver understanding, acceptance, comfort and ultimately harms safety.

My goal is to establish a novel, informative, cooperative and mutually adaptive interaction between driver and highly-automated vehicle. I will animate the currently inflexible haptic shared control interface based on fundamental insight into driver’ situated adaptation. First, I will develop estimation techniques that allow the haptic controller to determine individual driver’s situated control behaviour in real-time, recognize conflicts through force interaction, and to adapt itself accordingly to match decision and control strategies as well as safety and comfort margins between driver and automation. Second, I will enable the automated vehicle to communicate its own varying abilities and intentions through haptic and visual interaction.

I will develop the interfaces and techniques in driving simulators, before validating them in automated vehicles provided by the Dutch Automated Vehicle Initiative and Nissan. This breakthrough in shaping human-automation interaction is essential for accelerating a responsible introduction of automation in realistic driving situations, and is extendable to aviation and (tele)robotics.

VIDI

automotive, VENI, haptic shared control, VIDI
Project Output:

    Abbink et al. (2018) – A Topology of Shared Control Systems

    Scholtens et al. (2018) – A new haptic shared controller reducing steering conflicts

    Kolekar et al. (2018) – Modeling intradriver steering variability on sensorimotor control

    Barendswaard et al. (2017) – A method to assess individualized driver models

    Petermeijer et al. (2015)

Involved Companies:
Acknowledgement:

This work is part of the VIDI research programme 14127, which is (partly) financed by the Netherlands Organisation for Scientific Research (NWO).

Media Attention & Outreach

2019 06 “De Wereld van Morgen” TV interview and live demo
2016 08 E-book “Leven met Robots” – podcast
2016 07 Kijk – interview
2016 07 NRC – interview “Horse-riding on the highway”
2016 06 Delta – interview
2016 01 “Kennis van Nu” –  
TV interview
2016 01 “Kennis van Nu” – online media interview
2015 12 “RTL-Z” – TV item
2015 12 Transport Thursday
2015 11 Gala van de Wetenschap
2015 11 Studium Generale

Created by The Laboratory Network