Tubot: Idea Presentation

Table of Content

Project Timeline: June 2026 ~ July 2026

Final Result / Demo Visual

Introduction

Project Context

Blank Image for Project Overview

TuBot was a three-week robotics service proposal developed for WCRC 2025, the World Creative Robot Contest. The competition focused on creative, feasible robot-based solutions to real-world problems, and our team proposed an AI-enabled mobile robot service for urban flood prevention.

The goal was to design a realistic AI-enabled mobile robot service to improve storm drain maintenance. Our concept used an autonomous patrol robot to inspect storm drain inlets, collect GPS-tagged condition photos, and send the results to a platform where maintenance staff could identify problem areas and prioritize cleaning or inspection work.

Project Goal

My Role

My role / contribution

I proposed the core concept and helped shape it into a user-centred robot-platform service. My contribution focused on connecting the real maintenance problem, user workflow, operational design, and feasibility constraints into a coherent proposal.

Challenge 1: the Real User and Workflow Behind Storm Drain Maintenance

Moving From a City Problem to a Specific User

This project targeted a real city infrastructure problem, so the user could not be invented or assumed. Before defining the system requirements, we needed to understand who actually managed storm drain maintenance, what workflow they followed, and what information would help them make better maintenance decisions.

Challenge / Problem Visual 1

Pinpointing Who Actually Owned the Problem

We knew the problem was related to stormwater drainage, but “the government” was too broad to design around. I researched government websites and organizational structures to identify which department was responsible for storm drain and curb inlet maintenance, then narrowed the user from a vague public agency to the people likely involved in inspection, cleaning, and maintenance decisions.

Challenge / Problem Visual 2

Validating the Workflow Through Direct Contact

After identifying the likely department, I contacted government office staff and asked how their current workflow operated. I asked whether maintenance was routine-based, complaint-based, or priority-based, what constraints they faced, and whether location-based drain condition data could help them decide where to inspect or clean first.

Challenge / Problem Visual 3

Turning User Research Into System Direction

That research helped turn the proposal from a general robot idea into a service concept based on real user needs. The system direction became clearer: the robot needed to collect GPS-tagged inspection data, capture photos of drain conditions, and send the results to a platform where maintenance staff could review problem areas and prioritize work.

Challenge / Problem Visual 3

Challenge 2: Turning a Robot Idea Into a Complete Service Ecosystem

Robot Alone Was Not the Solution

The initial concept was an autonomous robot that could patrol sidewalks and inspect storm drain inlets. However, the robot itself would not solve the maintenance problem unless the collected data could reach the right users, support their workflow, and help them decide where to take action.

Challenge / Problem Visual 1

Defining the System Around Multiple Users

We separated the system into distinct user needs rather than designing solely around the robot. Robot operators needed a way to monitor patrol status, handle exceptions, and manage operations. In contrast, government maintenance staff needed a platform to review drain conditions, compare locations, and prioritize cleaning or inspection work.

Challenge / Problem Visual 2

Designing the Operational Flow Around the Robot

To make the service realistic, we considered how the robot would be deployed and managed in practice. This included when patrols would run, who would assign patrol routes, how operators would monitor multiple robots, when remote intervention would be needed, and how blocked paths or abnormal situations should be handled.

Challenge / Problem Visual 3

Turning Inspection Results Into a Usable Platform

We designed the platform concept around the information maintenance staff would need after inspection. The system needed to show drain locations, condition photos, status levels, priority areas, and regional summaries so staff could turn field data into maintenance decisions.

Challenge / Problem Visual 3

Challenge 3: Evaluating Feasibility Through Practical Engineering Constraints

Moving From Concept to Practical Constraints

Because this project was judged on realism and feasibility, the proposal could not remain a creative concept. We treated cost, weight, power consumption, battery capacity, runtime, driving speed, and inspection coverage as design constraints because the robot had to operate for long patrol periods while carrying cameras, batteries, electronics, and communication hardware.

Challenge / Problem Visual 1

Estimating the Robot Around Practical Components

I helped develop the hardware specification and bill of materials for the proposed robot. We estimated the required number of cameras, battery capacity, internal electronics, total cost, weight, power consumption, expected runtime, and coverage distance to evaluate the proposal as a buildable system rather than an abstract idea.

Challenge / Problem Visual 3

Choosing Server-Side Analysis to Reduce Onboard Load

We also considered whether AI inspection should run directly on the robot or be handled through a central server. For the initial concept, sending captured images through an internet connection and analyzing them on the server was more practical because it reduced onboard compute requirements, hardware cost, and battery load.

Challenge / Problem Visual 3

Grounding the Proposal in Measurable Operating Assumptions

The final concept estimated a robot cost of approximately CAD $2,500, a weight of 23.05 kg, a power consumption of 75.5 W, a battery capacity of 1080 Wh, a runtime of over 14 hours, and coverage of about 71.5 km per run at around 5 km/h. These estimates helped show how the system could operate under realistic cost, power, and deployment constraints.

Challenge / Problem Visual 3

Conclusion

Outcome

Our team developed TuBot into a structured robotics service proposal for storm drain inspection and maintenance prioritization. The proposal won the Gyeonggi Province Governor’s Award, receiving the Silver Prize (2nd Place) in the AI Driving Robot University/General Division at WCRC 2025.

Outcome Vissual

Refection

This project became a useful exercise in connecting robotics technology with real-world service value. Instead of treating the robot as the final solution, the proposal had to consider the actual user, maintenance workflow, operation model, platform design, and practical constraints such as cost, power, runtime, and deployment feasibility.

Extra Visual