Carmera coffee chat

Ever thought about how driverless cars would survive in NYC? We met up with Jan (‘18 BSE Mechanical and Aerospace Engineering) and Aravind (‘18 AB Mathematics) to hear about how they are working with Carmera (the name says it all) to make a driverless car-centric NYC possible.  

 Carmera is a company that currently uses their technology to provide architects, engineers, and stores with street IQ (like google streetview but updated everyday). Their long term goal is to create an autonomous car pipeline so that computers in autonomous cars will have up-to-date information (such as traffic or roadblocks) in order to plan the perfect route. Their technology could eventually take over the businesses of all navigation systems.

 

What are you working on?

Jan: I use landmark referencing for location determination. What that means is that I use parking signs, machine learning, and computer vision to ensure very accurate location coordinating.

 Aravind: I am trying to figure out how Carmera can use video feed from a phone camera to get simultaneous localization and mapping (SLAM). Usually, GPS provides very precise global location but it fails in the cities and is very inaccurate because of buildings and what not. That’s why I am trying to use the video feed from the smartphone camera to reconstruct the local map to enhance the location accuracy. This is a fairly recent field of research, thus I am reading a lot of research and looking at different implementations of SLAM and trying to find the best algorithms that work with the data Carmera has.

 

What does your average day look like?

Jan: The Carmera environment functions more like a research group than a strict corporate setting. I arrive at ten and then code for eight hours, eating lunch while I code. Then at four we check in with the team and ask other people if we have any questions and everyone brainstorms and gives us feedback. We have complete authority over our projects, they formulated goals for us and we reach out when we need help, but they let us go about it in our own way.  

 

How does it feel to have their trust and the responsibility of your own project?

Aravind: It is really nice to have no restriction on how you go about working. It is a lot of experimenting, sometimes successfully and other times not so successfully.

 

Have you had any a-ha moments?

Jan: One time I did not know what a particular piece of what I was working on was, so I asked one of the other team members about it. He looked at it for thirty minutes, after I had been working on it for a week, and it turned out that the thing I did not understand was actually holding everything for my solution. I did not previously know that what he showed me existed.

 

How do you fit into the team and what they are working on?

Aravind: Right now most of the team is working on the product that they just released. Many of them are focused on what sort of additions they can make. While they are focused on immediate expansions of their new product, we are working on enhancements that will be implemented in the future when they take their product to the next level and move on their plan to use the street IQ for autonomous vehicles.  

 

What are you hoping to take away from your experience?

Jan: I want to go to graduate school and focus on intelligent robotics in engineering, so what they are working on is exactly what I want to be doing. I have spent time understanding neural networks and computer vision at school, but I have not been involved in the programming side of things like I am now.  

 Aravind: I would like to go to graduate school, but I want to work in the industry for a few years so that I know what I want to get into. I do see myself working in this field and for a company like Carmera, I really enjoy the problem solving side of it.

Madelynn Prendergast