Drone Research; Palestinian Death

^{figures (a) and (b) from the GRASP Lab’s “SEER: Safe Efficient Exploration for Aerial Robots using Learning to Predict Information Gain”: arXiv:2209.11034}

by Penn Against the Occupation, 2025

Penn Engineering prides itself on an interdisciplinary, humanitarian approach to engineering for the betterment of the world. However, just a few layers of digging show a concerning level of genocidal complicity and militarized application to the department’s work. Professor Vijay Kumar is the dean of Penn Engineering and the head of Penn’s GRASP Lab (General Robotics, Automation, Sensing, and Perception Lab), and his face is often displayed smiling and speaking on display screens in Amy Gutmann Hall, Towne, and Levine. We took a brief look into some of the recent projects coming out of GRASP and followed their funding trails. 

A paper published by Kumar’s group in ICRA (International Conference on Robotics and Automation) 2023 entitled “SEER: Safe Efficient Exploration for Aerial Robots using Learning to Predict Information Gain” reads in the funding acknowledgments: “We gratefully acknowledge the support of ARL DCIST CRA W911NF17-2-0181, … NVIDIA, Israel Minister of Defense, …” among other grants, referencing the Israeli Ministry of Defense – not only a foreign defense ministry, but one of an illegal occupation force actively committing genocide. When professors at other schools have been confronted with taking research funding from the Israeli Ministry of Defense, a common deflection is that a student on the project is Israeli, and therefore was funded by their government’s department of defense similar to American students on US DOD-funded projects. However, something notable about the SEER paper is that none of the seven students listed are of Israeli nationality – of the list, international students are from China and Argentina, and yet IMoD is listed as supporting the project. The paper itself describes a learned framework for aerial quadrotors to navigate and fly through unknown indoor environments, specifically highlighting performance on a “SWaP constrained platform” – a term common in military research for technology that is limited in size, weight, and power by the conditions of war or battle.

The Israeli Ministry of Defense has a special interest in autonomous drones that can follow, surveil, and fire at Palestinians, both in research and in application (see documented IMoD quadrotor use for auditory terror, home invasion, and luring civilians out to fire on them, and for use as sniper drones). It is not difficult to imagine how a project like SEER, which involves autonomous methods of aerial entry, mapping, and indoor navigation on resource-constrained hardware, could fit into IMoD’s goals and interests.

This is not the only paper from Kumar’s group that works with the Israeli military. In a 2023 paper titled “Learning to Explore Indoor Environments using Autonomous Micro Aerial Vehicles,” authorship is shared between Kumar’s group and the labs of Ariel Barel from Technion-Israel Institute of Technology and Oren Gal from the University of Haifa. It is important that this paper is understood as more than a simple collaboration with Israeli researchers: since 2012, Oren Gal has been the CTO of the Autonomy and Data Science branch of the Israeli DDR&D (Directorate of Defense, Research, and Development) which is a joint government body between the Israeli Ministry of Defense and the Israel Defense Force (IDF), and he is also the former founder and CTO of the Autonomy and Data Science R&D branch of the Israeli Navy. Similar to the ICRA 2023 paper, the work focuses on performance of SWaP-constrained quadrotors for exploration and navigation of indoor spaces, emphasizing in militarized language that “SWaP constraints induce limits on mission time requiring efficiency in exploration.” 

The work on autonomous drone navigation coming from Kumar’s group is not fundamental science removed from military applications. These collaborations with the Israeli Ministry of Defense, either directly with research leaders or by receiving funding support, implicate both GRASP and the university as a whole as complicit in the ongoing genocide and illegal occupation of Palestine.

The interested actors in these projects also begin to reveal the underlying nature of much of GRASP’s research. Listed on both of these papers in the funding acknowledgment is US DOD grant ID #W911NF1720181, which is titled “ARCHES: Autonomous Resilient Cognitive Heterogeneous Swarms.” As of August 2025, ARCHES is currently allocated $43.1 million in total DOD funding, a number that has grown steadily via grant revisions from the initial allocation of $15 million (and has grown by $7 million in 2025 alone). Under the funding of this grant lives the DCIST CRA project, or the Distributed and Collaborative Intelligent Systems and Technology Collaborative Research Alliance. DCIST CRA consists of professors from 6 institutions: Penn, MIT, Georgia Tech, University of Southern California, UC Berkeley, and UC San Diego. Regarding Penn’s involvement, Vijay Kumar is the program manager of the entire consortium, with Alejandro Ribeiro as deputy PM and Profs. Dani Bassett, Kostas Daniilidis, George Pappas, and Ani Hsieh also committed to the project. Led by the US Army, the goal of DCIST/ARCHES is to create highly autonomous soldier drone swarms, which can act in parallel to and in collaboration with soldiers in “complex and contested environments.” Where exactly this technology might be deployed is also not hard to imagine: in the unclassified introductory slides to the project, an image of the streets of Homs, Syria is shown overlaid with images of American soldiers, drone fleets, and tanks (slide 5). Project descriptions submitted to the grant office describe enabling “force multiplication, the ability of a team of autonomous, heterogeneous mobile agents to greatly extend the capabilities and reach of soldiers, in yet-to-be-imagined ways.” 

Looking at some of the other papers from GRASP that are funded on this grant, we see:

• “Path Defense in Dynamic Defender-Attacker Blotto Games (dDAB) with Limited Information” (Pappas, Kumar): “… recent work has shown the effectiveness of these systems in related areas such as perimeter defense and surveillance.” 
• “Robust Multi-Robot Active Target Tracking Against Sensing and Communication Attacks” (Kumar): “Typical examples include … Adversarial agent tracking: Detect and follow adversarial agents that try to escape in an urban environment.” 
• “Vision-based Perimeter Defense via Multiview Pose Estimation” (Kumar, Jayaraman)
• “Adaptive and Risk-Aware Target Tracking for Robot Teams with Heterogeneous Sensors” (Kumar, Ribeiro, Yim): “…a team of robots with heterogeneous sensors must track a set of targets or hazards which may induce sensory failures on the robots.” 

Robotics research happening right across the hall is strengthening the fire force of American and Zionist imperialist war crimes across the Middle East and beyond. In 2022, GRASP Prof. Dan Koditschek held a workshop at ICRA 2022 titled “Workshop on Lethal Autonomous Weapons,” meant to discuss and debate the ethics of giving autonomous robotic systems lethal force. Although the writeup from the workshop is hosted on GRASP’s website (with the recordings from the panel found here), we are unfortunately past the point of debating when the imperialist bourgeois of the world stage (the US, Europe, Israel, and their proxies) hold an international monopoly on military robotics and have demonstrated time and time again that drones and legged robots are used to execute civilians at point blank range, police criminalized populations, and squash resistance in the face of genocide. There isn’t much space for ambiguity when a lab presents new drone navigation algorithms to an occupying military force that uses drones to fire on aid sites.

Whether the work is being funded by and reported to the US or to Israel, these projects have clear applications in armed military and surveillance settings in collaboration with governments that have brought military terror upon decolonial struggles for self-determination across the globe. The nature of the work, and the ways in which GRASP professors are marketing their research to garner defense funding, is an indelible moral stain on the university and all who work on these projects, unwittingly or not.  

Vijay Kumar and Penn GRASP Lab are guilty of aiding and abetting genocide and systemic imperial terror, whether these algorithms are deployed tomorrow or a few years from now. As students, workers, and researchers of this university, it is our moral duty to refuse consent to this complicity. Where we cannot comply in good conscience, we must speak, and where we are prevented from speaking, we must disrupt.

A Recent History of Campus Organizing | Home | Who Owns Penn?