Skip to content
OVER Business

Humanoid robots for retail, malls and commercial buildings

Humanoid robots for retail, malls and commercial buildings: after closing a robot walks the galleries, car parks and service corridors, checks exits and extinguishers and reports anything out of place. By day it gives spoken notices in public areas. Orders are plain language; the AI runs on your server.

Concept video

Indoor halls by day, perimeter fence by night

One robot covers a busy hall and, after closing, the fence line outside.

What do humanoid robots for retail and commercial buildings do?

In malls, offices and hotels, humanoid and quadruped robots take the rounds staff repeat every day: after-hours checks of galleries, car parks and service corridors, exits and extinguishers, and anything left where it should not be. They give spoken notices, and report each finding with a photo and its position on the map.

  • After-hours rounds. Galleries, car parks and service corridors checked after closing.
  • Public areas by day. Spoken notices and a visible presence in halls and entrances.
  • Safety checks. Exits, extinguishers and emergency routes present and clear.
  • Questions to the robot. “Have you seen a pushchair left in the car park?”, answered from its site memory.
Concept video

Autonomous perimeter round along a fence line

A quadruped robot follows a fence line at dusk and checks it sector by sector.

How do service robots find their way in a multi-floor mall?

They use the same visual positioning that OVER runs for visitors’ phones in shopping malls. The building is captured once with a 360° camera into a 3D map, and the robot localizes on it with its own camera. One operator mapped Città Fiera, 150,000 m², in 90 minutes. No beacons are installed.

  • One 3D map, many uses. The map behind AR wayfinding for visitors also positions the robots.
  • No-go zones. Shops, kitchens or staff-only areas kept out of every route.
  • Taught places. Name a spot once, “the north entrance”, and send the robot there by name.

Under the hood

One platform for every robot on the site

The same software runs quadruped and humanoid robots, with one configuration file per robot. A language and vision model turns the operator’s order into robot actions; a rule-based safety layer checks each action before it starts.

  • Orders in plain language. An operator writes “do the round of area B” or “go to the main gate” in chat, in English or Italian. The system splits it into robot actions and reports the outcome.
  • One web dashboard. Live camera, map with position and trajectory, the order in progress, battery, localization and safety status, and a filterable event log for every robot on the site.
  • Positioning without GPS. The robot localizes on the 3D map of the site with OVER’s visual positioning (VPS). Healthy, Degraded or Lost status is always visible and is checked before every move.
  • Rounds and patrols. To a point, along a sequence of waypoints or in a repeated patrol, avoiding obstacles. No-go zones drawn on the map are respected.
  • Open-vocabulary detection. The robot finds what you name, such as people, extinguishers, ladders, open doors or equipment, and can keep watch while it moves: on a match it stops and alerts.
  • Evidence and site memory. Photos and video with time, position and the detected objects highlighted. The system remembers places, sightings and tasks: “where did you see the extinguisher?”

Can a shopping mall robot work safely among visitors?

Every move is checked by a deterministic safety layer before it starts, and actions the rules flag go to a person. The robot avoids obstacles, keeps out of no-go zones and stops on its own if the link drops or the battery is critical. Staff can stop it from the dashboard or with “stop” in chat.

  • Evidence with expiry. Photos of people are kept for a period you set, then deleted.
  • Data on your premises. Camera images are processed on your server, not sent to a cloud.
  • Voice. Spoken notices in public areas through text-to-speech.

Safety

The AI proposes, the rules decide

The model never moves the robot directly. Every action passes checks that do not depend on the model, and people keep the last word.

  • A deterministic check before every action. Localization, navigation state, emergency stop, battery and allowed map points are checked by rules, not by the AI model. Each action is allowed, refused or sent to a person for approval.
  • Automatic reflexes. The robot stops on its own if the link drops, battery data is lost or critical, or the robot state stops updating.
  • Emergency stop. Always available on the dashboard and independent from the model. Only a named person can release it.
  • “Stop” in chat. A stop order written while the robot moves is recognized in about 0.1 s and cancels the motion.

How to start

One pilot site, then all the others

  1. 01

    Site survey and 3D capture

    We walk the site with a 360° camera and build the 3D map the robot localizes on. We check network coverage (Wi-Fi or 4G/5G), ground and lighting.

  2. 02

    One pilot site

    Robots, dashboard and the on-site server run your rounds for about 90 days, with results measured on your site.

  3. 03

    Roll out to more sites

    Add robots and sites with the same software and the same dashboard, without redesigning anything.

Robots in commercial buildings: questions operators ask

Can the robot use the map we use for visitor wayfinding?

Yes. The 3D map captured for AR wayfinding, the web map and kiosks is the same map the robot localizes on, so the building is mapped once for visitors and robots.

Can the robot guide visitors to a shop?

The robot walks to any point on the map and gives spoken notices, so it can accompany a visitor to a destination. Visitors can also follow AR directions on their own phones, with no app to download.

Humanoid or quadruped for a commercial building?

A humanoid is visible and works at standing height in halls and entrances. A quadruped is faster on long rounds, ramps and car parks. Both run on the same software and dashboard.

Does the robot need a cloud service?

No. Language and perception models run on a GPU server on your site, with no mandatory cloud service, so video and data stay on your premises. Robots connect to it over the site network through ROS 2.

Which robots does the platform run on?

Today a Unitree Go2 quadruped and a Booster T1 humanoid. The same software drives different robots through one configuration file each, and a simulator lets us test missions without the robot.

How does a pilot start?

With a site survey and a 360° capture of the area, then about 90 days on one pilot site with your own rounds. Tell us the site, the area and what the robot should do, and we scope it with you.

Project scoping

Plan a robot pilot on your site

Tell us where the robots would work and what they should do. A few clicks, then your contact details. Rough answers are fine.

Step 1 of 2

I’m…
What are you looking for? · choose all that apply
Where would the robots work? (optional) · choose all that apply
What should they do? (optional) · choose all that apply
Timeline and role (optional)
Timeline (optional)
Your role (optional)
Partner type
How would you like to work with us? · choose all that apply
Verticals you serve (optional)
Region, opportunities and team size (optional)
Region (optional)
Open opportunities right now (optional)
Team size (optional)

We reply by email within 1 business day to set up a short scoping call.

What happens next

  1. 01Answer a few questions. Two short steps. Rough answers are fine.
  2. 02Our reply. We reply by email within 1 business day to set up a short scoping call.