Jarvis Schultz

Jarvis Schultz's picture

Associate Director, MS in Robotics
Lecturer in Mechanical Engineering

Tech B290

+1 (847) 491-5474
Year graduated: 
Former Position/Degree: 
Ph. D.
A rendering of the robotic "puppeteer" I designed
for the robotic marionette project.

My PhD work was initially inspired by the robotic marionette project, where I worked on developing and experimentally implementing computationally efficient algorithms for automatically synthesizing controllers for desired marionette motions. The marionettes are extremely difficult to control; they are underactuated, highly articulated, nonlinear, and they possess closed kinematic chains. These complexities mean that even developing a software package capable of reliable, stable forward simulation proves to be quite challenging. Eventually we turned to the so-called variational integrators provided by the field of discrete mechanics. The variational integrators possess several key numerical properties that made them suitable for this application --- they have stable energy behavior, they exactly satisfy holonomic constraints, and they require minimal "user-tuning" to achieve reliable simulations.

When creating a variational integrator, there are several places one has a choice about which numerical approximations to utilize. My recent work has centered around a particular set of choices that yield a variational integrator that is well-suited to use in standard discrete-time optimal control and estimation routines. Specifically, I've been working on demonstrating and understanding the positive effects that this variational integrator has on these standard techniques. My ultimate goal is leverage these numerical advantages to facilitate high-performance control using inexpensive, low-performance infrastructure.

Google Scholar profile.


Northwestern University
Ph.D. in Mechanical Engineering, 2014
Dissertation: Discrete Mechanics Compuation for Real-Time Embedded Control

University of Wyoming
M.S. in Mechanical Engineering, 2009
Thesis: Meso-scale and Multicontinuum Modeling of a Triaxial Braided Textile Composite 

University of Wyoming
B.S. in Mechanical Engineering, magna cum laude, 2007

Northwestern University


University of Wyoming
  • ES 2410: Mechanics of Materials (Supplementary Instructor)
  • Engineering Summer Program: Introduction to Composite Materials (Instructor)
  • ME 2020: Intro to the Design of Experiments (Teaching Assistant)


I am an advisor the the Illinois Gamma chapter of Tau Beta Pi, the engineering honor's society.


I personally am an active user of the following software:

  • trep - dynamic simulation and optimal control
  • ROS - the Robot Operating System
  • GNU Emacs - an extensible, customizable text editor — and more
  • Python - a powerful, cross-platform programming language with many useful (and free) extension e.g. (NumPy, SciPy, matplotlib, and IPython)
  • LaTeX - the best way to write technical papers
  • Git - modern, distributed version control

Other Webpages