ICRA 2022 Plenary and Keynote Speakers

Ayanna Howard, M.j.t. Smith, Julie Shah, Antonio Bicchi, Joshua B. Tenenbaum, Marco Hutter, Michael Yu Wang, Veronica Santos, Professor, DrICRA 2022 · DOI

Abstract

People tend to overtrust sophisticated robotic devices, especially those powered by AI. As these systems become more fully interactive with humans during the performance of day-to-day activities, ethical considerations in deploying these systems must be more carefully investigated. Bias, for example, has often been encoded in and can manifest itself through AI algorithms, which humans then take guidance from, resulting in the phenomenon of excessive trust. This paper summarizes key insights from ICRA 2022, focusing on mitigating these ethical conundrums while advancing the state of autonomous systems, bio-inspired hardware, and human-centric AI.

Introduction

The 2022 IEEE International Conference on Robotics and Automation (ICRA) highlighted a pivotal shift: robots are no longer just isolated tools but are becoming integrated partners. This evolution brings new socio-technical hurdles. Researchers are now addressing the "Trust and Bias Gap," the economic impact of collaborative robots, and the "Embodied Intelligence Aporia"—the challenge of translating biological fluid intelligence into rigid mechanical systems.

Ethical AI and the Trust Phenomenon

A major focus of the conference was the risk of "overtrust." Dr. Ayanna Howard addressed how humans often over-rely on AI, potentially amplifying encoded racial and gender biases. These systems learn by mimicking human thinking, inheriting implicit biases that can impact high-stakes scenarios like healthcare and job placement.

Dr. Ayanna Howard, who presented on the ethical challenges of trust and bias in AI-driven robotics.

Advancing Cognitive and Physical Autonomy

To move toward more human-like intelligence, researchers are exploring new computational and physical paradigms.

Probabilistic Programming and Learning

Dr. Joshua Tenenbaum discussed the use of "probabilistic programs" built on physics engines. This approach seeks to combine built-in structural representations with data-driven learning, mimicking how human infants acquire knowledge about the world.

Dr. Joshua Tenenbaum, whose work focuses on building AI that learns and thinks like a human.

Bio-inspired Locomotion and Sensing

The field is also seeing a revolution in physical interaction. Reinforcement Learning (RL) has enabled quadrupedal robots to navigate complex subterranean environments, while bio-inspired hardware—such as gecko-inspired adhesives and optical tactile sensors—is giving machines a more nuanced sense of touch.

Real-World Scaling and Results

The transition from lab to field is well underway:

  • Autonomous Logistics: Zipline now powers over 70,000 km of autonomous flight daily, providing instant logistics to remote facilities.
  • Planetary Exploration: NASA’s Perseverance and Ingenuity are utilizing advanced autonomous navigation and aerial flight on Mars, surpassing the capabilities of all previous planetary robots.
  • Industrial Legged Robots: Legged systems have moved from research curiosities to established industrial products, proving their worth in challenging environments like the DARPA SubT challenge.

Conclusion and Future Outlook

The future of robotics is defined by integration. This requires scaling cognitive models to handle real-world complexity and ensuring that robotics knowledge is shared globally to avoid a knowledge skew toward developed nations. Most importantly, as robots and humans begin to share physical and cognitive spaces more fluidly, the development of ethical frameworks to mitigate bias and manage trust will be essential for improving overall quality of life.

Summary updated

Abstract

People tend to overtrust sophisticated robotic devices, especially those powered by AI. As these systems become more fully interactive with humans during the performance of day-to-day activities, ethical considerations in deploying these systems must be more carefully investigated. Bias, for example, has often been encoded in and can manifest itself through AI algorithms, which humans then take guidance from, resulting in the phenomenon of excessive trust. Bias further impacts this potential risk for trust, or overtrust, in that these systems are learning by mimicking our own thinking processes, inheriting our own implicit gender and racial biases, for example. These types of human-A! feedback loops may consequently have a direct impact on the overall quality of the interaction between humans and machines, whether the interaction is in the domains of healthcare, job-placement, or other high-impact life scenarios. In this talk, we will discuss these types of ethical conundrums in robotics and possible ways to mitigate its impact on our quality of life.
BibTeX
@inproceedings{howard2022icra,
  author = {Ayanna Howard and M.J.T. Smith and Julie Shah and Antonio Bicchi and Joshua B. Tenenbaum and Marco Hutter and Michael Yu Wang and Veronica Santos and Veronica Santos and Professor and Dr and Santos},
  title = {{ICRA 2022 Plenary and Keynote Speakers}},
  booktitle = {IEEE International Conference on Robotics and Automation},
  year = {2022},
  doi = {10.1109/icra46639.2022.9811971},
}
Figure