Profile

Cover photo
Ross Taylor
Works at Quanergy
Attended Louisiana Tech University
Lives in Houston, TX
953 followers|1,050,043 views
AboutPostsCollections

Stream

Ross Taylor

Shared publicly  - 
 
Algorithm competition once all vehicles are autonomous

This is an interesting discussion of what happens when the global optimum of travel is reached. Basically, the idea is that we tend to optimize for ourselves by getting newer and better technology. The author (and I) uses the Waze app to get an advantage on other drivers who don't. But when we are all in autonomous cars, companies that want to give an advantage to their customers will have to revert to tactics that harm others for the sake of the one.

Obviously, this will likely not be a significant problem for a long time and the global gains are worth any inequality for now. However, it will be interesting to see if this is an inequality that people are willing to accept (like airport fast lanes).
The wired metropolis creates clear winners and losers. The challenge ahead: Making the Smart City also an Equitable one.
1
Add a comment...

Ross Taylor

Shared publicly  - 
 
Baidu pushing infrastructure and behavior changes for vehicles

Baidu has published their vision for getting autonomous cars on the road soon. I'm a believer in the idea that full autonomy will be unlikely without infrastructure changes and that select routes will likely be where we see autonomous cars first. I don't agree with everything in this article but it is interesting nonetheless.
We should make modest changes to our infrastructure, emphasize predictable behavior, and teach the public new ways to interact with autonomous vehicles.
1
1
Add a comment...

Ross Taylor

Shared publicly  - 
 
Boston Dynamics humanoid progress

Boston Dynamics continues to make some impressive progress. The upper body is still pretty bulky but they've really trimmed things down overall. 
20
72
Antonio Pretto Melen!ez's profile photoJaveon Reeves's profile photoJohn B's profile photoRoss Taylor's profile photo
7 comments
 
+John B Stability. Although Boston Dynamics' robots have much less dependence on static stability than most humanoids, they still lean towards a fairly stable gait.
Add a comment...

Ross Taylor

Shared publicly  - 
 
Here's a (very) brief highlight from the +SpaceX​ landing +Yonatan Zunger​ shared earlier.
2
Frank Graffagnino's profile photo
 
Makes me wonder if in the future someone will build an island that is more in the launch trajectory for the first stage to return to. Would also be safer since you don't have to approach populated area. 
Add a comment...

Ross Taylor

Shared publicly  - 
 
 
The Washington Post: The $75,000 problem for self-driving cars is going away https://www.washingtonpost.com/news/innovations/wp/2015/12/04/the-75000-problem-for-self-driving-cars-is-going-away/
Giving a car "eyes" once cost a fortune. Now it's affordable, a good sign for autonomous vehicles.
View original post
6
Sandor Rezmann's profile photo
 
Cool!!
Add a comment...

Ross Taylor

Shared publicly  - 
 
New zero-g exercise machine

This device is a spin-off of the motor technology created for Robonaut.

A good friend is also spinning this off for commercial use: http://astrofittrainer.com/
1
Add a comment...

Ross Taylor

Shared publicly  - 
 
Boston Dynamics for sale

This is really too bad because the company was doing find on its own then Google bought it, took away all of the military contracts and is now abandoning it. I can't imagine who would buy it at this point because making a commercial case will be difficult and reestablishing the military relationship may be as well. 
The video, published to YouTube on Feb. 23, was awe-inspiring and scary. A two-legged humanoid robot trudges through the snow, somehow maintaining its balance. Another robot with two arms and pads for hands crouches down and lifts a brown box and delicately places it on a shelf -- then somehow stays upright while a human tries to push it over with a hockey stick. A third robot topples over and clambers back to its feet with ease.
1
Michael Interbartolo's profile photoRoss Taylor's profile photo
2 comments
 
We'll see. I could see Toyota because they do a lot of forward looking robotics research but I'd be very surprised by Amazon.
Add a comment...

Ross Taylor

Shared publicly  - 
 
GM buying Cruise Automation

GM is buying a head start on autonomous cars in the form of cruise. This is an interesting purchase clearly aimed at acquiring the team rather than the product. They seem to be planning to combine this acquisition with their investment in Lyft to compete directly with Uber and their attempts to replace drivers with self-driving cars (http://jalopnik.com/gm-and-lyfts-first-self-driving-cars-will-have-human-ch-1765117267). 
1
Add a comment...

Ross Taylor

Shared publicly  - 
 
First self-driving car accident in which Google is claiming some responsibility

This is a big deal. It was bound to happen. Assumptions are made all the time about how other drivers will behave on the road and those assumptions are bound to be wrong sometimes. In this case, the autonomous car and the bus driver did not make the same assumptions and a fender-bender occurred.

There is actually an argument that this is good rather than bad. These cars are not going to be perfect and it is actually a good thing for people to start getting use to the idea that there will be accidents. As long as the accidents are minor, it could lead to greater acceptance.
Well, it finally happened. One of Google’s autonomous vehicles might have caused a minor fender-bender. (Luckily, it appears nobody was hurt.) And guess what—it’s probably going to happen again. And that’s fine.
1
5
Add a comment...

Ross Taylor

Shared publicly  - 
 
 
Fast Company selects Quanergy as one of The World's Top 10 Most Innovative Automotive Companies in 2016

What gives autonomous vehicles their independence is the laser system that scans a car's surroundings for obstacles. Known as LiDAR systems, traditional setups are unsightly, require moving parts that sweep the entire 360 degrees around the car, and aren't particularly good at identifying specific objects, especially in bad weather. On top of all that, they're expensive: the device atop Google's self-driving cars costs upwards of $70,000. That high cost creates a barrier to speedy adoption of self-driving cars. In 2015, sensing tech company Quanergy Systems hinted at a solution: a solid-state LiDAR system that's the size of a credit card, costs just $250, and is more accurate than current options. "Our solid state LiDAR is significantly more compact and more affordable than any other unit on the market today, and will revolutionize the way automotive manufacturers design cars," said Louay Eldada, Quanergy CEO in a press release. The new system was unveiled in early 2016 at the Consumer Electronics Show (CES) and the company has already partnered with automakers including Mercedes-Benz, Hyundai-Kia, and Renault-Nissan.

http://www.fastcompany.com/most-innovative-companies/sectors/automotive

http://www.fastcompany.com/company/quanergy

View original post
1
Susan Taylor's profile photo
 
Congratulations. Great article.
Add a comment...

Ross Taylor

Shared publicly  - 
 
Nice summary of the +SpaceX​ accomplishment and its significance from +Yonatan Zunger​.
 
Huge kudos to +SpaceX for the successful launch and landing of a Falcon 9 rocket today. This is the first time anyone has ever flown a rocket to orbit and landed it upright. That's a big deal, because:

- Right now, rockets get dumped into the ocean or otherwise discarded after use, which is a big part of why flying to space is so expensive. It's like building a 747 and using it once.

- While people have flown to orbit and landed again (in the Space Shuttle), that required the use of a winged lifting body, which is enormously more complicated and expensive than a simple rocket. Also, the Shuttle required a big external tank and two booster rockets – which got jettisoned.

- While people have launched rockets and landed them vertically before (Blue Origin, just a few months ago), those rockets weren't capable of going into orbit. And going to orbit is a big difference from going into space: space isn't very far away at all, you just need to go about 60 miles straight up. Orbit isn't high, it's fast: orbit basically means that you're going fast enough that you keep falling towards the Earth and missing. Getting to that speed is a lot harder than just going up and falling down again.

- While people have tried to launch non-winged rockets into orbit and land them again (SpaceX, twice in the past year or so), it hasn't worked, because you're basically trying to land a pencil on its tip as it descends at a speed of a few thousand miles per hour. Previous attempts tried to land on boats (a pair of robot-controlled "autonomous spaceport drone ships" named the Just Read the Instructions and the Of Course I Still Love You), and the reason for the failures can be summarized as "now try to do that in high seas." This time, they landed on solid ground, which isn't as good from an orbital trajectories perspective, but which has the advantage of not moving about underneath you. (Usually)

This was a full-scale mission: a Falcon 9 took off from Cape Canaveral carrying eleven communications satellites. The first stage separated from the upper part of the rocket and landed safely, while the upper stages deployed the satellites into orbit.

So that's a major success for one of the most exciting companies in space travel today, and something that's likely to seriously cut the price of space travel over the next few years.
SpaceX’s Falcon 9 rocket successfully landed upright on solid ground at Cape Canaveral, Florida this evening, after traveling into space and back. It's the first time SpaceX has been able to gently...
58 comments on original post
3
Add a comment...
Ross's Collections
People
Have him in circles
953 people
Filipe Vilanculo's profile photo
Andrew Montpool's profile photo
Gary Downing's profile photo
Brandon Deitz's profile photo
Ewig Tour's profile photo
Simon Pollock's profile photo
Ines Sereno's profile photo
Superior Academic Writing Service's profile photo
Basilio Rosa's profile photo
Collections Ross is following
Education
  • Louisiana Tech University
    BS Mechanical Engineering, 1997 - 2001
  • University of Texas at Austin
    MS Mechanical Engineering, 2001 - 2004
Basic Information
Gender
Male
Story
Tagline
Roboticist and Computer Vision Engineer
Introduction
I'm a robotics engineer and a computer vision expert, an early adopter, and a technophile. These are the things I will post about most but I'll occasionally throw in some space stuff and other randomness.

I have two mechanical engineering degrees but I spend 95% of my time developing algorithms and software focusing on computer vision and robotics. I am also very interested in artificial intelligence but I'm still waiting for an opportunity to really delve into that field.
Bragging rights
I taught Robonaut how to move and took 3D pictures of the Space Shuttle in orbit.
Work
Occupation
Lidar, Computer Vision, and Robotics
Skills
Senior engineer with extensive experience solving complex problems in robotic control and 3D computer vision. Proficient in C++ development including object oriented programming, template programming, and code optimization. Thrive in a team environment and in customer facing roles.
Employment
  • Quanergy
    Principal Engineer, 2014 - present
    Working with Quanergy's Lidar sensors to develop smart sensing solutions for real-time 3D mapping and object detection, tracking, and classification. These technologies enable next generation driver assistance systems and autonomous vehicles.
  • Aerodyne Industries
    Robotics Engineer, 2012 - 2014
    - Developed algorithms for planning, control, testing, and operation of NASA’s second generation Robonaut humanoid robot. - Focused on the mid-level control framework including kinematics, dynamics, and trajectory planning as well as the high-level application development. - Developed with C++, Python, ROS, and OROCOS. - Developed a collection of tools to convert natural language project requirements into formal logic and verify the controller design. - Developed a model definition schema, data structure, and parser that allows easy access to important model parameters by the various parts of the software.
  • Neptec USA
    R&D Engineer, 2004 - 2012
    Managed small teams conducting 3D machine vision research, algorithm development, and software development for space, defense, and commercial applications. Projects included 3D vision system processing for navigation and surface inspection for a lunar prospecting rover. Developed algorithms and software for 3D machine vision data processing for space, defense, and commercial applications. Projects included ground and aerial LIDAR processing for object recognition and aerial LIDAR processing for removal of all non-ground points.
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
Houston, TX
Previously
Austin, TX - Ruston, LA - Shreveport, LA - Oak Ridge, TN