Basic Info

I am a postdoctoral research assistant at Robotics and Biology Laboratory in Technical University of Berlin. Previously, I earned a Ph.D. in Robotics from from Georgia Institute of Technology and finished my Bachelor's degree in the School of Computer Science, at Carnegie Mellon University.

I am interested in robotic grasping and manipulation; specifically, in the exploitation of soft materials and contact to simplify perception and control models for real-world scenarios. My work consists of two stages. First, I analyze experimental data from human subjects on how they grasp objects with both their natural and soft robotic hands. The goal of this analysis is to reveal the underlying patterns in their grasp strategies such as how they contact the environment, how their coordinate their hand and arm movements, etc. Second, building upon these insights, I develop motion planning algorithms that exploit contact with the environment to grasp objects under visual and actuation uncertainty.

  • Resume/CV - You can obtain my resume as a .pdf file here.

Soft manipulation

  • Human grasping analysis: Coordination of Intrinsic and Extrinsic Degrees of Freedom in Soft Robotic Grasping. Traditional grasp planners compute fixed contact points on an object and follow a simple coordination pattern for a hand's intrinsic (e.g. posture) and extrinsic (e.g. pose) degrees of freedom: the hand approaches the scene, its fingers close, and then it retracts. However, tightly coordinated intrinsic/extrinsic movements would enable a robot to adapt to the dynamic interactions with an object and its environment, and thereby exploit them towards successful grasps. As a first step in investigating coordinated behavior, we analyze how humans grasp objects using a soft robotic hand and show that increased coordination opportunities lead to higher grasp performance. Secondly, we implement one of the observed behavior patterns in a robotic platform and demonstrate an increase in grasp capabilities. Finally, we present results on how the exploitation of the interactions with the environment relax the constraints on scheduling intrinsic and extrinsic movements.

  • Robotic grasping: Please see here for my work on robotic grasping using environmental constraint exploitation and contact-exploiting motion planning.

Thesis work: MacGyver Project

  • Planning in Constraint Space: ICRA 2013: The goal is to autonomously design structures that are functional. For instance, what can a robot do if it needs to climb a tall structure or cross over a body of water and it can manipulate some objects? Moreover, how can we guarantee that the planner will find a solution if there is one? Here is an example where an agent comes up with such a plan where it uses a counter weight to build a bridge over some fire and climbs a height. You can try the code here.

  • Incorporating Kinodynamic Limitations in Simple Machine Designs: IROS 2014: Robots are inherently limited by constraints on their motor power, battery life, and structural rigidity. Using simple machines and exploiting their mechanical advantage can significantly increase the breadth of a robot’s capabilities. In this work, we present an autonomous planner which allows a robot to determine how arbitrary rigid objects in its environment can be utilized in machine designs to overcome physical challenges. We demonstrate results with Golem Krang, using levers as simple machines, to overturn 100 kg load and to push 240 kg obstacle.

  • Autonomous Realization of Simple Machines: ISER 2014: In this work, we report results on the autonomous realization of such a design by the humanoid robot Golem Krang, focusing on the challenges of autonomous perception, manipulation and control. Note that this is a long video - 4 min!. I'll re-edit it at some point :-).

  • Golem Krang - A Segway-like balancing robot which weighs more than 140 kgs (300 lbs) and can reach a height of 1.90m (6' 2''). I have been leading the development of perception, control and planning modules on the platform, and below are the technical reports we have written so far. The picture on the left is showing Mike Stilman doing push-ups on Krang while it is maintaining its balance by leaning backwards!! :-)

Additional Projects in Manipulation

  • Dynamic manipulation: Imagine a robot throwing an object across the room instead of navigating there among obstacles and placing it slowly. Or imagine an obstacle in a search and rescue scenario too heavy to move but the robot can push a bookcase onto it to knock it down. We are interested in exploiting the inertial properties of objects to extend robots' manipulation capabilities. To this end, we have written a number of reports, such as Closed-Form Rigid Body Orientations that explains how to efficiently use objects' inertial properties. Throwing and Sliding Rigid Bodies presents the math behind the following video.

  • Symmetry grasping: A large portion of everyday objects contain circular or mirror symmetries and in this work, we attempt to exploit this property by observing that any viable grasp at a symmetric part can be replicated around that part without additional computation (ignoring collisions between the rest of the manipulator and the world). This suggests, for instance for sequential task and motion planning problems, that we can find a single grasp that satisfies all the task requirements efficiently as the video below shows. You can find a report of the ongoing work here.

Technical Reports

  • M. Zafar, C. Erdogan and M. Stilman. Towards Stable Balancing. Georgia Insitute of Technology. GT-GOLEM-2014-004. 2014

Bloggity Blog

  • Useful Ubuntu Commands - This is the list of some useful commands that I keep forgetting.
  • Possible Mini Projects - This is the list of some projects that I want to complete either for educational reasons or completing previous work.
  • Random Code - This is a collection of random code I have written over the years.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License