Advanced motion control and GPS guide vehicle steering robot

A steering robot has been developed that uses advanced motion control and a GPS system to consistently perform vehicle assessment tests. Dean Palmer reports In the world of testing passenger vehicles, the test drivers cannot steer identically for each individual assessment, especially if the car's suspension has been changed between tests or if two different types of vehicle are being compared. However, this problem has been solved following the development of a steering robot system that uses advanced motion control integrated with a GPS system to guide the vehicle. The problem was first brought to vehicle test equipment specialist Anthony Best Dynamics (ABD) by Ford Motor Company, who needed to address the development of a repeatable 'roll over' test demanded by the US authorities. This imperative meant that ABD's engineers had to work hard and fast to develop the steering robot solution. So, a robot was developed by ABD in collaboration with motion control specialist Delta Tau. The robot is used to steer consistently through tests (including the roll over test) that determines a car's stability and safety when avoiding a collision. The steering robot clamps onto existing steering wheels or replaces the steering wheel entirely. Delta Tau's UMAC positioning system controls an electric motor that is used to actuate predefined patterns in the car's steering system. However, despite this, there are other factors that also affect the way the vehicle handles, such as wind, road camber and tyre friction. To overcome these, Ford and ABD used Delta Tau's motion control hardware and software, and went on to develop a 'path following system' that uses position feedback to ensure that the car follows the same route. Measuring the car's position on the road is achieved using an inertial and GPS navigation system from Oxford Technical Solutions. The path following system can be used to learn and accurately replay a course on the test track. It can also be programmed to allow repeated lap testing. Once the human driver has driven the course and positional data is collected via the GPS, it is processed within the UMAC controller. System accuracy is very high. In a recent test over one-and-a-half miles, the maximum path error was just 10cm and that was recorded during a 0.8G bend in the course. Typically, the path error was just 2cm. Even where there was a larger deviation of 10cm (due to high lateral acceleration) this was found to be highly repeatable and so the result was still the same for each lap. At the heart of the robot system is a Delta Tau UMAC position controller with a dedicated real-time processor. This controller can handle multiple motors and simultaneously run the 'path-following' control algorithm. And, because the UMAC made setting up the motors easy, it also enabled the development of the path-following programme to be completed quickly. A key factor in ABD adopting Delta Tau equipment was its ability to handle lots of data quickly, while also handling the motion control. The capabilities of the UMAC enable information to be captured from the GPS system via Ethernet and then processed. The position controller, which is usually mounted in the boot of the car, handles over 80 channels of data at a rate of 100Hz - giving 8kHz capacity, which is adequate to process the GPS feedback and the mass of data collected during each steering test. The position controller is connected to a PC through a USB interface, with communications to the GPS Motion Pack via an additional Ethernet interface. The operator steers up until the point where the robot takes over, but retains control of the brake and accelerator pedals. For constant steering pattern tests the operator holds a joystick; if the operator releases the joystick then the robot turns off immediately. This is an essential safety mechanism so that the operator can take over control of the car if the robot fails. A master power button is provided in case the software fails. A further benefit of UMAC is its ability to allow the steering robot to run autonomously, so that brake- and path-following robots can be added (also autonomously) but also co-ordinated when required within the whole system. The path-following steering robot has been used extensively for J-tests, Elk Tests, constant radius tests and, more recently, for repeating ride and handling tests on 'twisty' roads. Future steering robot developments from ABD are likely to address the trend towards testing differential braking systems. No doubt, as drive-by-wire technology becomes more commonly applied to steering systems, so too will the need to test it empirically emerge.