One Map
Deploy a Hundred Robots at Once
GUMA Robots · Shared Spatial Cognition Platform · Plug-and-play · Zero-code throughout
for fleet-scale operations
new robots ready on power-up
and fill in forms
live interface switch
· Cloud · Cross-border
Distribute One Map to Many Robots
Stop Letting Every Unit Run It Alone
Having dozens or hundreds of robots each re-build the map and re-home every day is a huge waste of labor. GUMA Studio lets the map one robot carefully built (point cloud + waypoints + semantics + no-go zones) be pushed to every robot on site with one click. It relays through the server: offline units queue up and get the update automatically when they come online — no map is ever lost. Check a newly purchased robot into the fleet and it can go to work the same day.
Command in Plain Language
No Coordinates, No Code
Commanding robots shouldn't be an engineer's privilege. Type "take me to the front desk" or "where's the fire extinguisher" and the platform locates it, plans the route automatically, and the robot walks there itself; when a visitor asks "where's the restroom", the robot leads the way. Draw a box around an area and name it "sofa area" or "main stage", and from then on one sentence sends a robot there — expertise settles into reusable semantic assets on the map for long-term reuse.
Dispatch Robots Like Playing a Game
If You Can Play, You Can Use It
Warcraft, Honor of Kings, Plants vs. Zombies — if you can play these, you can use GUMA Studio. Ctrl+F1/F2/F3 forms teams and picks units; a single hotkey selects a robot instantly. WASD moves forward/back/left/right, Q/E turns in place — hold to move, release to stop, safe by nature; right-click the ground for "navigate here". Take over manually anytime, and the red emergency stop at the top halts all navigation and mapping with one press. Scheduled tasks even let robots clock in on time and return to charge on low battery — fully unattended.
Skills That Talk and Perform
10 Agents · 100 Skills
Attach "hear a word, do an action / announce something" skills to a role: say "do a dance" and it dances, say "welcome" and it waves automatically, hit a keyword during a talk and it gestures automatically. The dance library includes kung fu, popping, ballet and Wing Chun — previewable and choreographable into shows; hosting/announcing lets you audition voice timbres; photography shoots in 1080P HD; VIP face recognition for exclusive reception; teaching interaction actively finds people; shopping-guide and tour waypoints carry narration. Guide, tour, host and photo roles all set up in a few steps, then sync across robots once activated and saved.
Fleet-wide Data Dashboard
The More They Run, the Richer Your Enterprise Assets Grow
See every robot's battery, memory, uptime, health score, photos and videos on one screen. Search and sort to instantly spot the one about to run out of power. The system auto-classifies status into three alert tiers — critical / warning / notice — so you don't have to stare at the screen. One sentence gets AI to produce a fleet operations analysis report. Photos and recordings captured on patrol settle into the media library automatically — the longer you use it, the thicker your maps, semantic areas, detection tags and visual assets become.
Your Toughest Pain Points, All Handled by One Platform
On the left is what's giving you a headache, in the middle is how GUMA Studio solves it, on the right is what you get.
*The above illustrates typical platform capabilities; specific features and deployment forms should be discussed with GUMA Robots based on your actual business scenario. Some advanced scenario modes and industry skill packs are continuously updated.
Beneath the 5 Capabilities · 8 Functional Modules at a Glance
Shared Spatial Cognition
One-click map distribution to many robots, indoor/outdoor mapping toggle, auto-relocalization on power-up, virtual fences (no-go / virtual wall / corridor / glass zones), configure a role once and bind many robots.
AI-Native Mapping
Robots auto-explore and map as they move, pick maps by quality, box and name semantic areas, one-sentence semantic navigation, draw obstacle-avoidance no-go zones, rename and reposition landmarks, one-click task execution.
Gamified Dispatch
Ctrl+F1/F2/F3 team select, WASD movement + Q/E turning, right-click-to-navigate, posture switching (stand / dance motion control / AI mode), one-click task dispatch, pause and emergency-stop anytime.
Agents & Skill Library
Ten scenario Agents (scheduled tasks / knowledge management / spatial semantics / world model / skill invocation, etc.) + a hundred Skills (hosting / dance / narration with gestures / photography / VIP reception / shopping-guide & tour).
Enterprise Knowledge & Roles
Documents (PDF/Word/TXT/MD) into a knowledge base, one-click knowledge base from a website URL, AI one-click persona expansion, voice-timbre auditioning, capability toggles, activate for formal duty.
Data Asset Management
Robot operation detail table (search/sort), long-term recognition tags, tiered fleet alerts, AI fleet analysis, media library that settles photos and recordings — the more robots run, the richer your assets.
Data Security & Deployment
Standalone / LAN / on-prem server / cloud server / cross-border server — five modes on one back office, with login gating + account-to-robot binding of visible robots for data isolation.
Advanced Skill Packs & Scenarios
Fixed-point patrol photography, live-stream selling and shopping guides, walk-and-narrate tours, greeting reception with area-based people-finding, import industry skill packs + long-term role memory; advanced scenario modes keep expanding.