TRIP-Bag: A Portable Teleoperation System for Plug-and-Play Robotic Arms and Leaders

Noboru MyersEqual Contribution, Sankalp YamsaniEqual Contribution, Obin KwonEqual Contribution, Joohyung Kim*
KIMLAB, University of Illinois Urbana-Champaign
first author denotation These authors equally contributed to this work. * Denotes corresponding author.
arXiv Video Playground

Abstract

Large scale, diverse demonstration data for manipulation tasks remains a major challenge in learning-based robot policies. Existing in-the-wild data collection approaches often rely on vision-based pose estimation of hand-held grippers or gloves, which introduces an embodiment gap between the collection platform and the target robot. Teleoperation systems eliminate the embodiment gap, but are typically impractical to deploy outside the laboratory environment. We propose TRIP-Bag (Teleoperation, Recording, Intelligence in a Portable Bag), a portable, puppeteer-style teleoperation system fully contained within a commercial suitcase, as a practical solution for collecting high-fidelity manipulation data across varied settings. With a setup time of under five minutes and direct joint-to-joint teleoperation, TRIP-Bag enables rapid and reliable data collection in any environment. We validated TRIP-Bag’s usability through experiments with non-expert users, showing that the system is intuitive and easy to operate. Furthermore, we confirmed the quality of the collected data by training benchmark manipulation policies, demonstrating its value as a practical resource for robot learning.

Video

Hardware Components

Teleoperation Pipeline

Within the commercial suitcase, an aluminum frame is used to reinforce the overall structure. The frame additionally secures all of the electronics and 3D printed mounts for the leaders and follower arms. The mounts are designed to be compatible with our custom 7-dof robotic arm, PAPRAS (Plug and Play Robotics Arm System), transmitting power and communication and enabling rapid deployment. Each arm holds an RGB-D camera at the wrist, with a third camera mounted on a pole to capture the full scene. The system utilizes external power, which enables long duration, stable data collection.

The main electronic components are a power supply, follower PC, leader PC, router, and power strip. In the fully packed state, TRIP-Bag weighs just under 30kg complying with typical airline overweight check-in baggage limits.

Teleoperation

Our overall teleoperation and data collection pipeline is based on the Plug-and-Play Robotic Limb Environment (PAPRLE). Please refer to the PAPRLE project page for more details.

Teleoperation Pipeline

How to Setup TRIP-Bag

Leveraging a commercially available suitcase, TRIP-Bag is designed for seamless transport and rapid deployment. From initial unpacking to the start of teleoperation, an expert user requires an average setup time of 3 minutes and 27 seconds. The setup procedure is shown below.

* The teleoperation nodes are launched from an off-screen operator

Where has TRIP-Bag Traveled?

With the use of a commercial suitcase, we are able to easily deploy the system at various environments. Explore the locations we have deployed TRIP-Bag: Click the buttons below to jump to a specific region, or click the pins on the map to view site-specific videos/photos.

Data Diversity

Data is only as strong as its variety. With the collection of our dataset we observed the unpredictability of human movement, accounting for unique collector attributes, collection postures, and environmental fluctuations. Explore the visual breakdown of this diversity below.

Single User Multiple Environments

Campus Restaurant
Main Quad
Dorm
Illini Union
Beckman Institute
House 1
House 2

Multiple Users Single Environment

User 1
User 2
User 3
User 4
User 5
User 6
User 7
User 8
User 9

Example Policy Rollout

Using TRIP-Bag, we collected a diverse dataset of 1238 demonstrations across two manipulation tasks: fruit collecting and egg cracking. We trained imitation learning policies on this dataset, and evaluated the policies on a physical robot arm. Below are example rollouts of the learned policies.

Task 1: Fruit Collecting

Task 2: Egg Cracking

Other Bimanual Tasks

To evaluate the versatility of TRIP-Bag, we deploy the system across a diverse set of manipulation tasks that stress different dimensions of teleoperation performance. In particular, we focus on scenarios that require long-horizon execution, high-precision manipulation, tight bimanual coordination, and high-payload handling. These tasks reflect common challenges in real-world manipulation, where robots must maintain stable control over extended interactions while coordinating multiple limbs under significant physical constraints.

Multi-Step Meal Preparation
USB Plug Insertion
Folding Shirt
5 lb Payload Transport