Segway Uses Software to Go from Innovation to Production
The Segway Human Transporter (HT) is the first of its kind—the first self-balancing, electric-powered transportation machine. With the ability to emulate human balance, Segway HT uses the same space as a pedestrian, and can go wherever a person can walk.
To move from concept to production, engineers at Segway LLC needed to overcome many engineering hurdles to produce a reliable, durable, and comfortable product that could be manufactured in quantity for a worldwide market. Due to the innovation of the product, Segway needed versatile software to solve unique engineering challenges and, ultimately, ensure that the Segway HT would meet its lifecycle goals, reliability standards, and performance metrics.
At the core of the Segway HT's dynamic stabilization, or self-balancing technology, is its inertial balancing system, which was modeled and verified using Matlab (technical computing software) and Simulink (modeling/simulation software) from The Mathworks. Using Matlab, engineers were able to run extensive tests to verify specific performance and durability goals for the product, as well as its mechanical integrity. As part of product development, Segway's engineering team relied heavily on the software's modeling capabilities to visualize various performance aspects of the propulsion system. With the ability to model the battery, motor, and actuators, engineers were able to understand the electromechanical dynamics of the propulsion system and accurately identify design challenges early in the testing phase, prior to production.
To solve the problems of battery and capacitor life, the Segway battery system* was modeled extensively and experimentally verified using Matlab. Numerous tests were run to determine the drain on the battery and electrical components at various speeds and terrain slopes. Segway engineer J.D. Heinzmann wrote a custom M-file to graph the results to help visualize the impact of various design alternatives. His work resulted in changes to the motor drive control algorithm, reducing current ripple and increasing capacitor life. These routines were also used to analyze energy consumption and identify opportunities for increasing the usable range of the HT.
To meet unit cost goals, Segway HT uses a cost-effective microprocessor to balance itself and the user, and to manage overall operation of the machine. This microprocessor has fixed-point mathematics capability and a limited amount of memory, necessitating efficient, compact implementations in the HT's software. The motor drive development team used Matlab and Simulink to analyze various polynomial approximations to fixed point trigonometric functions. This ensured that the mathematics would work correctly when implemented on the Segway HT microprocessor.
Segway's mechanical integrity engineers also used Matlab to perform fatigue analysis and ensure the Segway HT will meet five-year design life. Fatigue analysis data sets were very large, containing 96 minutes of live ride data. The software enabled engineer Andrew Steiner to confirm that the life of the HT's component parts would fit within the design goals. Engineers could also plot all of the data, automatically pull out the top 10 data values, and plot the details around each of these 10 values. As a result, Segway was able to identify potential failure modes and correct design problems before going to production.
Using the software, the Segway engineering team designed signal processing and control algorithms, evaluated the dynamic and mechanical performance of the Segway HT, and created algorithms for fault detection and response.
* Segway LLC, in cooperation with the U.S. Consumer Product Safety Commission, has conducted a voluntary recall of Human Transporters in 2003 to correct a hazard that can allow the rider to fall under some operating conditions, especially if the HT's batteries are near the end of charge. A free software upgrade is available from the company to remedy the condition. For more information on the recall, click here .
For more information, visit www.segway.com and www.mathworks.com .