Redesigning Simulation Mapping for Iterative Workflows
The Challenge: RealSimE operates in the autonomous driving simulation market, where AD/ADAS engineers need high-fidelity HD maps to test their systems in virtual environments like CARLA, dSPACE, and RoadRunner.
When I joined, RealSimE was an MVP with no persistence, an unclear export flow (no token/data preview), limited region tools, and no in-app simulation preview for validating OpenDRIVE. With October 2025 trials only 12 weeks away, we risked breaking real workflows, losing early enterprise trust, and delaying GeoMate’s launch in a competitive AD simulation market.
View live product
The impact
3 Biggest Pain Points
No persistent, reusable simulation map workspace
Every new scenario required reloading GIS layers, re-selecting AOIs, and re-applying rules
Engineers struggled to track versions and recreate past experiments
Heavy back-and-forth with GIS/mapping teams
Coordinates, shapefiles, screenshots, and notes were scattered across email threads
Small geometry or rule changes took days to align
OpenDRIVE validation required constant tool switching
Engineers juggled QGIS, CARLA, OpenDRIVE viewers, internal scripts, and logs
Small geometry or rule changes took days to align
They had to mentally stitch together geometry, topology, and schema correctness
Problem Statement
“Autonomous driving engineers need to efficiently create simulation maps, persist and manage multiple versions, and validate them without tool-switching or GIS back-and-forth.“
User Flow Optimization
Before jumping into the final design solutions, I first mapped out the existing user flow of our old demo. By doing this, I could clearly see where the major issues and gaps were. In that initial user flow, several pain points stood out:
First, users had to reselect the same area every time they logged in, which meant no work persistence and no way to organize multiple areas or scenarios in one map.
Second, the download progress was essentially fake, and users weren’t informed about their token usage or how much data they had left.
Third, drawing boundaries was limited to a simple box, making it hard to select specific streets. Finally, the project section used latitude and longitude instead of recognizable names, and users had no clear feedback about token consumption.
From Initial Pain Points to New Solutions
After identifying these issues, I created a new optimized user flow to address each problem. Now, we have a persistent workspace, clearer token visibility, more flexible drawing tools, and a more intuitive project organization. With these improvements, we’ve streamlined the entire process before diving into the design solution.
Design Solutions
Iteration & Refinement
Maze usability testing
After completing Version 1.0, I conducted Maze usability testing with 25 internal beta users to validate the redesigned workflow. Testing revealed two critical friction points requiring rapid iteration:
