Automated assistance in construction sites 

Productivity in construction is lower than other industries

 The construction industry has suffered from an immense productivity challenge for the last three decades That’s because construction sites are complex environments that require more than just management and software tools for productivity gains, which is what also makes them huge opportunities for hardware startups in the years ahead.

Material management

What happens if rovers can solve the problem of tracking and deliver goods in construction sites saving time and money?

Why rovers?
More accurate, more payload and better energy consumption performance

High rise building


In construction sites is possible to find two different systems to deploy materials and connect different levels in high rise buildings. We are focusing our project on lift connection with storage in key floors.

Service structure

Idea is to connect tools and materials using rovers. Vending machines provides materials and tools according with every trade, using as a reference the most common elements they need during the work.

Using a mobile app workers can ask to rover materials they need, rover connect with vending machine carring the tool. At the end of the path, rover find the worker using navigation and gps. When reach the point, release the tool using face recognition as a security layer.

Market survey

Asking to potential users in the market, they answer that last time purchases increases their budgets, they also needs 1 hour or more to go outside to purchase. It happens very often during the time lapse of the construction site.

Market analysis

Structure  analysis

To analyze work patterns in construction site, we divided it in two areas According to site logistics we can track crew, tools and materials

Flow chart  showing  robotic path from point A (tool) to point B (worker)

Focusing on navigation we can find two types


To run a navigation demo, we mounted a system with Turtlebot 2 robot, Astra Camera, ROS under Arduino 16 and RVIZ to visualize maps

Demo 1 – Autonomous Navigation

Using  system proposed ,  robot did an autonomous path from point A to point B using navigation packages in ROS

Robot: Kobuki Turtlebot 2| Location: IAAC | Controller: Auto| Camera: Orbbec Astra|Platform: ROS Kinetic |Connection: Laptop – Kobuki

Demo  2: Autonomous navigation with obstacle avoidance

Robot: Kobuki Turtlebot 2 |Location: IAAC Controller: Human |Camera: Orbbec Astra |Platform: ROS Kinetic|Connection: Laptop – Kobuki| Obstacle: Human

Demo 3  – Labour Following navigation

Robot: Kobuki Turtlebot 2 | Location: IAAC | Controller: Human | Camera: Orbbec Astra | Platform: ROS Kinetic | Connection: Laptop – Kobuki | Obstacle: Human

Demo 4  – Labour Following Navigation (Carrying Plier)

Robot: Kobuki Turtlebot 2  |Location: IAAC |Controller: Human| Camera: Orbbec Astra |  Platform: ROS Kinetic | Connection: Laptop – Kobuki |Obstacle: Human


Face recognition

Face recognition demo 

Using face recognition to activate hardware gripper using raspberry pi  and raspberry camera

System setup

 Future exploration 

 Tool detection

Dataset construction for handtools

Workers pattern analysis

Final idea