Stanford Artificial Intelligence Laboratory

Keeping with our laboratories tour through the world, now we present one of the laboratories involved in the Darpa Challenge Stanford Team, which was actually the winner or second place in the last 2 events.

One of the main figures of the laboratory is PhD Sebastian Thrun, who is one of the authors of the great book Probabilistic Robotics, and we have here in the laboratory our own Japanese Copy of the book.

Among some of the research of this lab is  A Robotically-Augmented Walker for Older Adults and a very interesting 3D mapping system using robots for small environments such as mines or caves.

I his page there is a number of videos we are sure you will enjoy watching.

Well, tomorrow we will report on the other members of the Stanford Darpa Challenge winners. And as well we will have our open laboratory so we will post some pictures and some insight as well.

Remember you can follow our updates in our Twitter Account

Research Group at Cornell University

In this our last post on mobile robotics, we are also beginning to introduce little by little laboratories that are located all over the world that are currently working in mobile robotics as well as some other interesting technologies.

Professor Mark Campbell's Research Group has as an objective the research of autonomous and semi autonomous mobile systems, such as mobile robots, flying robots, and the Darpa Challenge vehicle.

Part of professor Campbell staff was part in the Darpa Challenge Team of Cornell University which was in the few teams that where able to finish the course, which itself is already a great task.

Among their research topics range: Cooperative Tracking, Decision Modeling, Sensor Fusion, Large Scale Robotics and Human-Robotic Interaction.

We hope in the future we can keep reporting on news over this and other laboratories over robotics.

Remember you can follow our updates in our Twitter Account

Wolfram Alpha

Hello readers. This time we would like to introduce you to a most useful tool in the scientific field.

From the makers of Mathematica , Wolfram Research, we are presented with an innovative and quite good searching engine, which provides results unlike any search engine you have seen.

Lets say you are looking for the flight distance between Tokyo and Mexico City, lets input that in Google and you will get something like this, which is mainly composed of commercial information, which is really useful when trying to buy a ticket, yet in the scientific world most of the times we are interested in something more.

And thus, Wolfram Alpha search result is this, which if you are looking for geometric information or doing some research is way better.

It is not a better search engine, it is a different kind which aims to give more specific and exact results to more concise questions.

I leave you with Wolfram himself explaining the searching engine.

Remember you can follow our updates in our Twitter Account

Visitor from Boston University

Hello, we hope you have a nice week, and following up our posts on mobile robotics, here we present now, the work a PhD candidate at Boston University presented in our laboratory a week ago.

His name is Morteza Lahijanian and he is a member of Anderson Laboratory at BU. His paper was on Symbolic control of mobile robots, which is an area that deals with the common problem of mobile robot navigation and control, but using more an abstract concept like full commands instead of complex equations, into trying to conceptualize the entire system.

His paper name was Automatic deployment of autonomous cars in a robotic urban-like environment (RULE), and he showed us in his power point presentation how a simple set of commands could been mapped and assembled in accordance to the environment and user necessities.

For the moment we do not have a video, but as soon as I can find something nice, we will post it.

Have fun and have dreams.

Visitors from Berkeley

Last week we had the visit of  Haomiao Huang and Jeremy Gillula from Stanford University.

They presented their paper Hierarchical, hybrid architecture for robot motion planning and control. In which they describe a way to plan the motion and control of their mobile platform STARMAC.

The STARMAC is a flying vehicle equipped with sensors, that you will be able to see in action in the following video, it has been subject of some excitement in the scientific press world.

 

Remember you can follow our updates in our Twitter Account

Mobile Robotics in Michigan University

Remember this is the 3rd in a series of 5 posts on mobile robotics, so you are actualized with what the academic world is doing currently.

As we mentioned before, we had some visitors last week, one of them Vishnu Desaraju, who presented a very good paper on mobile robotics in the ICRA in Kobe. While in Michigan University, he worked in the Multi-agent Decision and Control Lab where they developed a robot collision avoidance system, work that he presented in the ICRA.

The slides of his presentation are here and the accompanying paper  in this link.

His work posses a very interesting approach on how autonomous vehicle can interact with each other without the necessity of GPS systems or very advanced localization systems, and excels in demonstrate how in a simple architecture one can achieve great degree of control.

Pictures and Movies can be seen in the Michigan University site that for the sake of avoiding right infringement we will not post.

Thank You Vishnu and we hope in some next posts, we will be able to post something on your lab at MIT.

As well we invite you to follow our lab posts and latest activities in our Twitter Account

Tsukuba Challenge

Hello again, and we hope you have the nicest Monday of all.

Today we are talking about the Tsukuba Challenge, this was inspired by Darpa Challenge,, but it highly differs of it. In that all the computational work has to be done in usually one laptop, since this challenge is oriented only in mobile robot navigation.

The Tsukuba Challenge was an organizational effort started by Prof. Yuta of Tsukuba University and Prof . Hashimoto form or very own laboratory. It has as an objective that an autonomous navigated robot should complete a given path.

To finish the route, the robots usually only carry one laptop and in it they have to do all the navigational process.

In 2007 the path consisted in a straight course with no return in a park in Tsukuba City, Tokyo. This year Course was completed by 3 teams, 2 from the Tsukuba University and 1 from The Kanazawa Institute of Technology.

In the 2008 the path became more complex, involving a loop circuit, this time only the Yamaha Corporation Robot was able to finish the route .

This year challenge consist in the robot having to follow a course that involve among other things a small park and a bridge over a lake.

We are going to put a video that has a part of the course that was defined for the 2008 Tsukuba Challenge.

Darpa Challenge

Hello Readers.

This time we are planning on updating you in the current status of autonomous navigation systems, this will be a series of 5 posts, being this first one the introduction of the Darpa Challenge.

The Darpa Challenge was created with the objective of developing the next generation technology for keeping the war fighters out of harms way, in order to that vehicles have to be intelligent and capable of self maneuvering along some predisposed terrain or some city.

First Year’s challenge was to move the vehicle across the Mojave Dessert in a given route. Second Year was to move it along streets, where the vehicle had to make turns, park and some other maneuvers, such as respect traffic signals and speed limits.

In the first competition, Carnegie Mellon University Robot was the number one with the most miles covered, without completing the route.

In the second Year the prize went to Stanford University, with Carnegie Mellon University coming in second place.

The last year was the Urban Challenge which took place in Victorville, California and the prize now went for Carnegie Mellon University.

This challenge has inspired much others all over the world, which we will be talking here in the future.

Until then we leave you with this video of the competition.

Again, the DARPA people is kind of jealous so here is the link

We keep having visitors

Yesterday we had a busy day with the visit of two American Universities, Gregory A. Cole from Worcester Polytechnic Institute in Massachusetts and MIT's Vishnu Desaraju.

Gregory presented a very nice introduction and explanation of a MRI Compatible robotic system, which we should be posting within the week when we can get gold of his information.

Vishnu presented some of his background as well as some introduction to his current research topic which has relation with autonomous flying mechanisms.

 

They both are presenting their papers in ICRA and we wish the the best of lucks.

Mid Term Presentation

Hello, this past Friday, Leon presented his midterm presentation titled A compressive sensing and swarm optimization algorithm for 4W1Hin the ISpace.

This is the paper along with the slides of the presentation.

We hope you enjoy it and present any ideas or suggestions in the topic.

Cornell University Visitors

Hello, we hope you have the nicest of the weeks, and here we are again, with current news of our Laboratory. Today we had the visit of Jonathan Schoenberg who presented his paper Localization with Multi-Modal Vision Measurements in Limited GPS Environments Using Gaussian Sum Filters.

He explained to us this was the algorithm they used for the DARPA Urban Challenge in which they were able to finish the competition.

We think this kind of experiences are very important to promote growing and to allow laboratories across the world to interchange experiences and points of view over different issues.

So for you to share some of the experience here is a video of the GTalk given by the DARPA Challenge Team in Google.

 Here is the direct link in case you cannot watch the embedded video.

 

 

I hope you can all watch the video, we had some problems with the embedding of this one.

Information on the go

Since the dawn of computing, one of the prime objectives has been making them accessible to everyone and to minimize the size as much as possible.

It has even existed the idea of computing without computers, in which we can actually have access to our programs, files and mails without a computer, nowadays cell phones gives us some of this functions, yet limited to their size and capabilities.

Sixth sense is a portable system that projects different kinds of information on any surface, making in that moment any surface a touch screen, for example you can project a keyboard and start writing, you can project a screen and read your mails.

It is developed by MIT and explained in this video.

Jazz and Robots

Now, have you ever wondered about imagination. That very thing that makes us better than automatons is that we can imagine and create solutions out of nowhere.

In the arts this is called improvisation, the art of one being able to create something out of thin air, just by following the flow of the situation.

Georgia Tech Center for music technology proves us that robots are no longer estranged with self creation, and gives us Shimon, a robot which can follow the lead in some music piece as shown in the video.

Can you imagine the implications of this.

Back from Golden Week

We just came back from golden week, to have our weekly meeting. Leon presented his Rinko presentation titled

A compressive sensing and swarm optimization algorithm for 4W1H
in the Intelligent Space

Abstract:
The 4W1H technique has been a subject of study in the ISpace, presenting multiple problems that had to be solved individually in order to improve performance and number of sensors. The current paper presents a proposed algorithm that makes use of sensing and processing tools compressive sensing and swarm optimization to pose a  feasible solution to this problem.  

Also Tamura san presented an interesting paper and Kou presented one more chapter of the book “The design of future things”