Technology 3 min read

MIT Team Develops Visual Navigation Tech For Delivery Robots

Andrey_Popov /

Andrey_Popov /

Initiatives across the world are already exploring the option of delivery robots. Machines can now bring your order straight to your doorstep – whether its a mail or a snack.

While aerial drones delivery is more popular, robots that are built to roam the city still have a place in automated delivery service.

Earlier in the year, Amazon launched a service that it’s referring to as Amazon Scout. It involves using what’s essentially a hamper on wheels to deliver drop off packages to customers.

Similarly, Starship Technologies announced a deal to deploy a fleet of 25 robots to deliver takeouts at George Mason University.

Here’s how the current delivery machines reach a customer’s door.

The Current Navigation Tech For Delivery Robots

These machines often navigate using a combination of GPS and the maps of a neighborhood or a whole city. Then operators have to enter a set of pre-loaded coordinates to enable the robots to reach their destination.

Since these machines only travel a short distance to deliver goods, the current navigation system might be an overkill. So, the researchers at MIT developed an alternative approach.

Instead of using a pre-loaded map, the new navigation technology could enable robots to find the door using visual cues.

A graduate student in MIT’s Department of Mechanical Engineering, Michael Everett said:

“We wouldn’t want to have to make a map of every building that we’d need to visit. With this technique, we hope to drop a robot at the end of any driveway and have it find a door.”

Here’s how it works.

How the Visual Navigation Tech for Delivery Robot Works

The new approach involves training machines to understand and label visual objects as humans do. For example, by recognizing a driveway or garage, the robot can make an educated guess on where to go.

To create the new tech, the MIT team used existing navigation algorithms that enable robots to identify visual objects in the environment. Then, they improved the algorithm to allow delivery robots to make decisions based on what they see.

Next, the researchers tested the technology in simulations using satellite imagery of neighborhoods. They also applied the algorithm to an unknown house outside of the training data.

Findings from the test suggest that the navigation algorithm can enable delivery robots to find the front door 189 percent faster than the current solutions.

Everett noted:

“Even if a robot is delivering a package to an environment, it’s never been to; there might be clues that will be the same as other places it’s seen. So the world may be laid out a little differently, but there are probably some things in common.”

As you may have guessed, the breakthrough could lead to more efficient delivery robots in the future.

Read More: Training Robots to Understand What Humans Want

First AI Web Content Optimization Platform Just for Writers

Found this article interesting?

Let Sumbo Bello know how much you appreciate this article by clicking the heart icon and by sharing this article on social media.

Profile Image

Sumbo Bello

Sumbo Bello is a creative writer who enjoys creating data-driven content for news sites. In his spare time, he plays basketball and listens to Coldplay.

Comments (0)
Most Recent most recent
share Scroll to top

Link Copied Successfully

Sign in

Sign in to access your personalized homepage, follow authors and topics you love, and clap for stories that matter to you.

Sign in with Google Sign in with Facebook

By using our site you agree to our privacy policy.