Wednesday, May 07, 2014

Wireless Sensor Network Made Easy

Wireless Sensor Network Made Easy

I have been working on wireless sensor network for the last few years. This includes student projects as well as industrial projects.

Basically, it involves having a suite of sensor nodes communication with each other and finally the entire suite of data collected by that sensors nodes can be presented to you. The collected data can further processed before a pictorial view of what make sense is presented.

I worked from ground up - that is the sensor node. Each sensor node comprises of the following:
a.     Embedded microprocessor
b.    Sensor and its suitable interface circuit to the microprocessor
c.     Radio for wireless data communication
d.     Battery or power source as each node should stand alone

As the data and information is cascaded upwards, it will involved the various protocols for data communication and how these data is transferred. It also involved getting the data to a cloud so that you can access the data from your smartphone.

I have the intention to create and made that sensor nodes as well as the various hardware look simple. A setup for wireless sensor network in term of data acquisition before data can be associated, analyse and annotated.

I called it Project WSensNet. More updates can be found at https://sites.google.com/site/wsensnet/


Friday, March 25, 2011

Line Tracing Robot

Just update a video and some pictures on this link https://sites.google.com/site/myiiprojects/

This projected completed recently by my students. The robot uses a simply web cam to detect and track a line.

The algorithm is build based on OpenCV under Visual Studio 2008 C# environment. Using only a web cam, the robot manage to track a line. The algorithm manages to filter off noise and unwanted false line detection to create a robust tracking. From the video you can observed that cracks on the floor does not affect the robot motion.

Tuesday, July 10, 2007

MozzieBot

A meeting with NEA, showed a need for a robot that can conduct search in places that the inspectors has problem acessing, and yet has to do so crawing in to conduct inspection because of the recent dengue fears.

Technology has progressed a lot. Video pipe back can easier be done by using IP camera and the viewing can be done using a wireless PDA. This made the "portable" needs smaller, as compared to any other handheld device. Imaging strapping the PDA to the back of your wraist, while holding on to the control device to control the robot.

Sunday, July 08, 2007

Video Filming for Project X - 8 July 2007

I was tasked to do filming. But the video is for someone else. I don't mind the filming - got a camera man to help me. What frustrate me is the re-shoot -due to the fact I do not know how to read the client mind.

The client does not specifically say how the screen shots to be like, just the general flow of the video. He did not state the background, the props, the angle of shot etc.

I just submit the video and hopefully, that's the end. Because of this meeting late into Thursday night, followed by a re-shoot on Saturday - nothing to gain. Feedup!

Monday, January 08, 2007

Busy Busy 2007

The year 2007 will be a busy year for me, especially in the area of engineering.

I am handling 4 projects, managing and executing them. All of them are robotics related.

From the oldest contract to the newest signed, they are:
a) Project A - converting two All terrain vehicle to drive-by-wire and added in vehicle control
b) Project L - Develop algorithm for a CMU robot and bench mark the performance of the results with the base-line codes
c) Project R - Develop a robotic platform with manipulator arm as payload, as well as integrate semi-autonomous function into the robot
d) Project S - Organise a robotic competition

A lot, a lot of tasks to perform and I pray tht as I multi-task between the projects, I will be able to engage God's help.

Friday, December 08, 2006

Drive by Wire Conversion

Not every vehicle comes with drive by wire.

During the recent Darpa Grand Challenge 2005, the Red Team from CMU and the team by Stanford U uses commercial available vehicle and converted them to drive by wire. The Red Team uses a HUMMV while the Standford uses a SUV.

I find it easier to get a vehicle off the shelf rather than develop one. There are a lot of good car manufacturers. Before selection consider the following:

Cost,
Space and payload
Speed
Controls available - Throttle, Brake, Steer and Brake.

I will talk more about these controls later.

About Me

As an engineer, I always gets questions on how I do this and how I do that.

I graduate with a post graduate degree in Control Engineering. Prior to that, I am trained as a Electrical and Electronics engineer in my first degree.

I have worked as an engineer for more than 10 years, working on various projects. During the recent years, my involvement is in the area of robotics, in particular, unmanned ground vehicles. It involve discipline from mechanical engineering, electrical engineering, software engineering, communication engineering, to research in control algorithms, navigation, vision, pose estimation, learning algorithms, sensor fusions etc.

I am fortunate to be able to work with research institutes and various industries on a number of such vehicles for the past 6 years.Currently, besides the technical lead, I also manage such projects. My two recent projects is on a unmanned Polaris All Terrain Vehicle, a 1-Tonne wheel vehicle, as well as developing learning algorithms for a robot developed by CMU. Another new project, which just began is to build a low cost 30kg autonomous robot.

I hope to share, as much as I can and without revealing "secrets" abouts these projects.