Starship Delivery recently deployed a fleet of robots to deliver food across the Ole Miss campus. The robot delivery service begins January 22. Photo by Billy Schuerman.

You’ve seen the robots, but how do they work?

Starship Delivery recently deployed a fleet of robots to deliver food across the Ole Miss campus. The robot delivery service begins January 22. Photo by Billy Schuerman.

The robots that have roamed campus for the past few weeks are officially beginning their food delivery service today. A lot of factors – keeping delivery costs down, making the robots travel efficiently and protecting them from the elements – went into creating the robots that now roam the Ole Miss campus.

Starship Technologies, the company that provides the delivery robots for the university, worked to make the robots cost-effective to cut food delivery costs. Light Detection and Ranging (LIDAR) sensors, typically used to maneuver autonomous robots, were out of the question.

Instead, the robots rely on neural networks to navigate the world. 

Starship’s neural networks rely on data gained by radars, cameras and ultrasonics. This data is then processed by the neural networks. Instead of programmers writing the code manually, the neural networks “learn” and write the code. 

When training the robots to travel, programmers must be careful to account for all situations. Potential issues include cars being spotted in mirrors or a robot seeing a motorcycle for the first time.

Architectural engineering can help with some of these issues. If the robot can put an object into its global context, it can determine if the path in front of it is a sidewalk or a road. 

Solving problems with the robots can be tricky. For one, the issues of space and cost restraint can cause engineering problems. Batteries, data storage and sensor calibration have to be carefully engineered to optimize performance and keep costs down. 

Bugs in the software can be extremely difficult to fix. In addition to the large amount of data involved in training the robots, there is the problem of deciding which solutions to implement. One solution could optimize the average performance of the robots, but it would decrease their ability to function in sunny weather. 

Despite the complexities of the robots, they have been successful in multiple cities and in multiple weather conditions – including snow, rain and ice. Right now, Ole Miss has robots all over campus. Today, you’re able to get deliveries from them.

Previous Story

Ole Miss basketball remains winless in SEC with 73-48 loss in Knoxville

Next Story

The robots are coming: automated food delivery begins on campus

Latest from Blog

US Air Force: Why It’s The Best

Lorem ipsum dolor sit amet, ei officiis assueverit pri, duo volumus commune molestiae ad, cum at clita latine. Tation nominavi quo id. An est possit adipiscing, error tation qualisque vel te. Stet

Margherita Pizza: The Recipe With Videos

Ius ea rebum nostrum offendit. Per in recusabo facilisis, est ei choro veritus gloriatur. Has ut dicant fuisset percipit. At usu iusto iisque mandamus, simul persius complectitur at sit, aliquam moderatius elaboraret

Jazz Music: Untold Stories

Labore nonumes te vel, vis id errem tantas tempor. Solet quidam salutatus at quo. Tantas comprehensam te sea, usu sanctus similique ei. Viderer admodum mea et, probo tantas alienum ne vim. Eos

How to become a successful blogger?

Quo natum nemore putant in, his te case habemus. Nulla detraxit explicari in vim. Id eam magna omnesque. Per cu dicat urbanitas, sit postulant disputationi ea. Duo ad graeci tamquam interesset, putant
Go toTop

Don't Miss

It’s all Greek to them: why some students leave their fraternities and sororities

Greek life is commonly associated with student life at the

Designing Your Oxford Bucket List (w/ DM A&C Editors Clay Hale & Caroline McCutchen)

Welcome back to the Weekly Scoop! In the first episode