Tuesday, April 1st, 2025
The Taub Faculty of Computer Science, Room 337
First part of the event included potential advisors that gave “lightning talks” about their research to promote their lab and recruit new graduate students from TASP. Second part included graduate students presenting posters about their ongoing research.
First place, best poster award: Sean Man
Second place, best poster award: Lilach Biton
Third place, best poster award: Rotem Kain
Master’s degree graduation ceremony, 2024
A special TASP seminar where potential advisors gave “lightning talks” about their research to promote their lab and recruit new graduate students from TASP.
Abstract :Recently we witnessed an increasing trend for autonomous aerial vehicles in the city’s sky to deliver goods, transport passengers, attend medical support and more. These vehicles perform high mobility for point-to-point missions, but to do so, they should cruise across the sky with minimum interference. As these trends grow, more vehicles will demand these safe routes as more potential collisions require resolution.
Here we give some insights to a preliminary novel concept of a low-altitude air transport system (LAAT) that supplies the infrastructure for the aerial vehicles to move in the urban area. Inspired by the well-researched ground transportation methods, the suggested LAAT system is based on the natural relation between density and accumulation of the vehicles in a given area – as described in the macroscopic fundamental diagram (MFD), and then using a flow control method called perimeter control to regulate the inter-regional flow from adjacent regions.
A micro-simulation of a LAAT vehicle was designed, including mission definition and collision avoidance mechanism, and then it was implemented in an aggregated simulation to show that an MFD-like relation exists in a given region of LAAT system similarly to ground traffic MFD. Then, an adaptive perimeter controller is designed and tested in several multi-regional simulations. The controller performance was measured in the meaning of network congestion avoidance in different scenarios, including different number of connected regions, different regional and inter-regional flow demand (attraction) and tested on the plant model – either original non-linear model or simplified partially-linearized model.
The simulations show that the aerial vehicles flow indeed perform a MFD relation, and the extracted MFD was used to design a perimeter controller. The simulations show that the controller successfully prevents congestion on the partially-linearized model of the multi-regional scenarios and extends the stability of the system to harsh demand in the original non-linear model of the multi-regional scenarios.
Afterwards, robustness research was performed on the designed perimeter controller to evaluate the dependency of the solutions improvement on several key parameters under expected real-life implementation difficulties such as controller-plant channel delay and control communication constraints.
Ronen Brafman – Ben-Gurion University
When: 08.06.2022 at 15:00
Where: zoom
Abstract: While one-trick robots are abundant in industry, constructing an autonomous robot that can perform multiple tasks flexibly, such as a service robot, is a difficult task, we are yet to reach. A large number of robotics researchers have been working hard to design and implement algorithms for diverse robotic skills. But these skills are hard to put together: A lot of software engineering and thought goes into each script that operates them. In this talk I will present a research program seeking to change this situation, providing plug’s play autonomy for robotic via what we call the autonomous operating system (AOS).
This system uses a language we designed for documenting robotic skill code to provide the AOS with the information needed to decide when and how tol execute each skill.
No additional software-engineering effort is required beyond this specification.
I will describe our initial version of this system and our future plans.
The talk is non-technical and requires little background.
You can see the seminar here
When: Wednesday, April 27th at 15:00
Where: Zoom
Abstract: In this talk, I first discuss how we leverage machine learning methods to generate cooperative policies for multi-robot systems. I describe how we use Graph Neural Networks (GNNs) to learn effective communication strategies for decentralized coordination. I then show how our GNN-based policy is able to achieve near-optimal performance across a variety of problems, at a fraction of the real-time computational cost. Finally, I present some pioneering real-robot experiments that demonstrate the transfer of our methods to the physical world.
When: Wednesday, April 13th at 15:00
Where: Zoom
Talk Abstract The questions of dexterity, agility, and learning from a few demonstrations have intrigued robotics researchers. In this talk, I will explore answers and solutions to these questions via the following case studies: (i) a dexterous manipulation system capable of re-orienting novel objects. (ii) a quadruped robot that is substantially more agile than its counterparts (runs, spins) on challenging natural terrains. (iii) framework for learning task-sensitive perceptual representations for planning and out-of-distribution generalization. While a lot of recent progress in robotics is driven by perception, we show that learned controllers can help address problems that were previously thought to be hard. I will discuss our findings, the insights we gained, and the road ahead.