Workshop Highlights | From Black Box to Open Book: Cracking the Code of Cognitive Robotics (ERF 2026 Workshop)

Our workshop took part in the early morning on Tuesday, March 24th, namely it was part of the first batch of ERF workshops. During the workshop, we had a combination of short talks followed by a two-part interactive session. The main highlights of the workshop are presented below in a visual summary.

Invited talks

We had four invited talks during the workshop. Prof. Javad Ghofrani gave a talk on the topic "Explainability, Cognition, and Embodied AI in HRC".

Dr. Alex Mitrevski gave a talk titled "Creating Interpretable Robots in the Connectionist Era: A Noble Goal or a Fool's Errand?".

Sarah Terreri presented PAL Robotics with the talk "Welcome to PAL Robotics, where the future is being reimagined every day".

Finally, Lakshmy Arvind presented the COGNIMAN EU project with the talk "Cognitive approaches for estimating challenging manufacturing processes".

Interactive session

During the interactive session, participants were invited to put their thoughts on various short questions, which resulted in word clouds related to each of those questions (the word clouds will be posted on this page soon). Questions from the participants were discussed as well. The interactive session was split into two parts. In the first part, the topic of robot explainability was discussed.

We specifically discussed the following questions:

Which best practices are currently regarded as most effective for ensuring explainability?
To what extent are they implemented in existing systems?
What forms and levels of explainability are expected by end users?
How do these expectations differ across user groups and usage contexts?

The word clouds of the participant responses are shown below:

The second part of the interactive session was dedicated to the topic of scalability of robotic solutions.

Here, we discussed the following questions:

What are the key barriers limiting robotics adaptation in regulated environments like medical devices?
What are the common technology gaps limiting robotics adaptation in regulated environments like medical devices?
What strategies can bridge the scalability gap between lab demos and real-world robotic deployments?

The word clouds of the participant responses are shown below: