Steve Norris - Software Engineer, Roboticist, Writer
I started in electronics in the early 1970’s. I took electronics at Nashoba Technical High School in Westford MA and worked part time for a company that was using a new device called a microprocessor. It was actually the Motorola 6800. From there I went to Wang Laboratories and worked from the board repair department all the up to the Artificial Intelligence development group working with MIT on model-based diagnostics. Today I work as a software engineer/architect for Phios Corporation producing custom applications in the health care sector.
In addition to my day job I also build unusual robots and then write about them for Robot Magazine.
For my robotics work, I prefer the Parallax Propeller chip as the main controller. The Propeller is ideal for robotic work and is in fact actually eight individual microcontrollers called Cogs built into one chip. This means you can have up to eight separate operations running at the same time performing sensor processing, motor control and RF communications. The Cogs can perform simultaneous tasks cooperatively or individually while sharing resources through a common hub.
For simpler tasks I prefer to use the Arduino Uno. It’s perfect for quick prototyping and applications.
As a software engineer I work in Microsoft C# using Visual Studio 2010. I also like working the new Microsoft functional language called F#. In some ways it reminds me of my old Lisp days. For my robotics work I use either the Parallax Propeller IDE to write code in Spin or in Propeller assembly language.
These days I do not have much on my bookshelf since I started using a Kindle. I do have the usual books on software development, robotics research as well as books that support my casual interest in physics.
There are two projects that stand out. The first is an interesting health care application used by insurance companies to issue individual health assessments. It is a web-based questionnaire that in turn will produce a detailed action plan to assist the member in adopting a healthier life style. I built the reasoning system based on an older AI technology called Case-Based Reasoning (CBR). This proved to be excellent solution to the underlying classification problem that’s needed to produce an action plan.
In the robotics world my most favorite project was the Beverage Delivery System. The Beverage Delivery System (BDS) is a system that implements a fully automated inter-home delivery system. Anyone who has built a robot has, at one point or another, been asked the question “but can it bring me a beer?” I’m sure that like me, roboticists all over the world have grown weary of this question. The Beverage Delivery System is the answer to this age old problem, a problem that has plagued mankind since the invention of the canned beverage.
The system uses a coordinated dual robot system. Baxter is the transport component and supporting Baxter is a fixed location dispenser robot called RoboFridge. You can think of a RoboFridge as a kind of soda vending machine that does not require money. Like a real vending machine RoboFridge supports the dispensing of canned beverages. The concept of the BDS could even be extended to non-food items like medications, DVDs and books.
The BDS process starts when a user requests an item using a simple remote control. Baxter is activated from his sleep mode and navigates to RoboFridge using a variety of navigation techniques. These include dead reckoning using encoders, IR beacons, and line following. Once he arrives Baxter will initiate a docking sequence with RoboFridge. After docking, RoboFridge deposits the canned beverage into Baxter’s cargo area. Baxter then undocks and navigates back to the thirsty user to deliver the beverage.
Finally I can get a frosty cold beverage delivered to me in my favorite easy chair without any effort or argument!
Certainly the greatest reward to all my work was being selected as one of the twelve inventors to appear in the National Geographic Channel series called Mad Scientists. Each episode of the series highlights the work of an individual maker/inventor. My show focused in on my collection of robots that I have built over the past several years for articles that have appeared in Robot Magazine. There is a “reality” spin to the show where the inventor is challenged to build something with the show’s host (John Bowler) in just two days. For the grand finale the “build” is revealed to an audience of friends and family and put to the test. For me the challenge was to build a robot that could play the basketball game of PIG. We were actually quite successful in building such a robot and it made for a great show.
For my current robot project I am working with the Microsoft Kinect. This sensor has certainly become all the rage in the robotics world. It has some pretty impressive depth sensing and mapping capabilities all for $150. The disadvantage is that need to be connected to a PC running Windows 7. I’m working on an article now on how to use it in robotic applications.