Steering you right?

Frank Massey continues his look at the thorny issue of autonomous control, focusing on steering and stability correction leading to autonomy

Published:  09 May, 2019

Picking up from my topic and opinions on autonomous vehicle control last month I think it reasonable to explore the very technology our safety is to be placed.

When considering the most challenging aspect of autonomous vehicle control, we must look to steering and stability correction. My references are limited to the Volkswagen Group, however most manufacturers now share similar drivetrain and chassis technology.

Steering assist
These systems have evolved over many years in what I term modular development. Steering assist is such a system. Steering assist is directly proportional to driver input force, the steering torque sensor g269 detects rotation, the steering angle sensor g85 provides angle and rotation acceleration.

Responding to this data the control module j500 calculates the required assistance from the power steering assist motor v187. When parking, a low or zero vehicle speed combined with a rapid steering input provides maximum assistance. During driving additional data relating to environmental conditions, urban or motorway, modifies the appropriate assistance.
One of the first problems to overcome was return to neutral or zero steering angle. This is activated with a reduction in force on the torsion bar, whereby the rate of return is also a function of environmental influences. The dual steering angle sensor is comprised of a LED and photo electric diode.

The steering torque sensor operates on the magnetic resistance principle. Failure results in a gradual reduction in assistance. The asynchronous brushless motor provides up to 4nm of assistance. Once again emphasis should be directed to programming and adaptive correction via e serial platform.

Stability & proximity
When introducing vehicle stability dynamics, even more data is required: An accelerometer as well as yaw and  pitch sensors will complement existing input requirements. Enhanced and shared functionality with ABS enables the braking system to support vehicle control through corners by applying a control force through the rear brakes.

We now need to consider the vehicle proximity control system; the system employs an ultrasonic sensor to monitor and determine the environment. However, this interim system has several critical shortcomings, especially due to its narrow detection field and inaccurate position calculation regarding other vehicles and obstacles. The next modular enhancement introduces side or blind spot monitoring or side assist. This system also has limitations with range and vector limitations. Although operating on a high speed can network, it operates on a master slave principle, for example; slave units only transmit data and diagnostics on demand from the master module.

It is of note that the vehicle now relies on no less than 13 control modules, with predictive position algorithms. Later evolution will include optical, video, ultrasound, infrared and laser. Optical lane assist which is mounted on the windscreen requires considerable coding and calibration, notwithstanding windshield replacement, so much for off-site repairers.

Calibration & correction
Calibration requires determination of the camera orientation, the exact installed location, the height at which the camera is installed and three orientation measurements. This is an electronic function as no mechanical adjustment is provided. Therefore any change in tyre, wheel diameter or suspension repair or modification will invalidate this system accuracy, including fault memory errors.

We now move into the era of de-coupling direct driver steering input. This system allows computer correction of steering angle. For example, with a loss of driver control, ESP can introduce a counter steer input to regain control. This system is intended to maintain the maximum static traction between the road surface and tyre. Should this be insufficient to maintain a safe curve radius, the ABS can be employed to help recover the vehicle attitude.

The system can carry out actual steering angle correction while the driver maintains a different steering wheel input, such as on snow, ice, or on flooded road surface conditions. In order to facilitate this function, a mechanical flexi-coupling is mounted in the upper steering column. The outer has 100 teeth, with the inner posessing 102. They can rotate together as one with direct driver input command or can rotate at a different angle disengaging driver direct input control.

In effect this system still complies with statutory requirements as having still a de-facto fail-safe mechanical connection between driver and steering mechanism. Therefore is still level 0 status, in terms of autonomy.

At this point we are a million miles from even level 2 or 3 autonomous control. Level 3 allows for the driver to release physical contact with vehicle controls yet remain available and alert in case of system failure. Please make your own mind up. However, I’m not for turning!



Related Articles

  • No self control? 

    Having witnessed the growth of passive driver assist systems and the intent to move towards fully autonomous vehicle control, my topic this month is to raise both thought and debate towards the implications. My first intention is to separate assistance from autonomy.

    I fully support assistance as it provides a safer environment for the driver to concentrate on vehicle control. Many of these systems have been available for a very long time, including possibly the very first, power steering and power windows.

    ABS to power steering
    Anti-lock braking systems (ABS) are, I think, an excellent example where drivers may be misled as to the safety improvements. However, the laws of physics still apply, and the co-efficient of friction and kinetic energy will always dictate the retardation distance and vector. Obvious enhancements to ABS work as a fully integrated system, including dynamic chassis stability.
    Early variants simply monitored the wheel speed sensor frequency, reducing the engine throttle angle to reduce torque through the driving wheels when a significant differential existed. Recent additions now include variable geometry anti-roll bar and adjustable rate shock absorber damping with self- levelling.

    Evolving in parallel with these systems, and this is where there is an arguable transition from passive to active or automatous control, is the steering system. The introduction of power steering does have great advantages in reducing driver fatigue and improving mechanical response to steering wheel input. The next evolution was variable rate steering assist, whereby the assistance is proportional to steering angle and road speed. with the evolution of brushless motors and highly accurate position sensor technology, steering systems now offer corrective suggestion to the driver via a subtle torsion bar within the upper steering column. Should the driver resist this small force the system will disengage leaving the driver fully in control.
    I am choosing to ignore for the moment fully autonomous steering control as it embodies a whole array of additional control input requirements.  This allows me to focus on some of the more peripheral driver support systems which I do fully endorse. Matrix vehicle lighting control is possibly one of the best safety improvements. This enables full beam lighting always, yet avoiding oncoming vehicle light stray. Smart cruise control is also especially useful on motorways in uniform traffic conditions.

    Compliance
    The next group of driver assist starts to cross the boundaries of assistance, this is due to the introduction of long- range transmitters and receptors, lane divergence, and vehicle proximity awareness. This technology does of course lend itself to other previously mentioned systems.  

    There should be a very sobering pause at this point.  To maintain system integrity and accuracy from the above systems a little thought should be given to the almost non- existent function called calibration.it is critical. If you fully consider the implications of everyday servicing and repairs that affect these systems, compliance is the responsibility of the repairer. This means you.
    This is the point where I cannot avoid the transition towards full driverless autonomous control. Due to several critical considerations, technical compliance, political compliance, legal compliance, and public acceptance, it is to be rolled out in five steps over several years. Ford recently suggested it could be implemented by 2021, with level zero full human control, to level five where the human has no input responsibility.

    What of the globe’s biggest commerce giant’s? Intel has just purchased an Israeli autonomy tech company for $15 billion. Google has spent a modest $30m, and Facebook is in it too. All hellbent on convincing us of the benefits in total vehicle automation. Given their past and current dishonesty, self-interest, and responsibility avoidance you can bet it all going to be a financial beartrap.
     However, my personal feelings are more complex. Humans has evolved over many thousands of years by overcoming and controlling a multitude of challenges. It has enabled our brain and cognitive functions to develop to incredible levels. Imagine then, being trapped in an autonomous container with absolutely no functional requirement. What will you do by way of brain stimulation or choices. I accept traffic jams are worst than toothache, but driving is a socially shared experience. Think of the simple activities that release endorphins, such as cycling and walking. Why? because of the brain stimulation and cognitive responses, a form of achievement.

    If you must have total autonomy for your travel requirements, then public transport is available now. My acid test for the techno maniacs out there is, given that the technology is currently available and has been proven over several years, would you choose to fly in an aircraft with no pilot? Remember that even in autopilot there are teams of humans constantly monitoring the flight path and technical systems.

    Credibility
    I’m not ignorant of the accident statistics that give credibility to automation, if that was the true motivation, then smoking and alcohol would be banned tomorrow as they kill and maim an awful lot more.

    It has been suggested that our home environment would be improved as our car could drop us off and then park its self in a less congested place, so if you live in central London your car could end up in a South Downs village. On a more sinister note, if an autonomous vehicle faced with an inevitable collision from a oncoming car, would it mount the pavement and choose the mother with a pram as the better survivable outcome for its occupants?

    The very best qualities of life always come back to interaction, be it with other people, pets or machines, what next? When do machines decide we are the redundant component? Disagree, or debate, but don’t accuse me of not embracing technology, I have spent my life trying to master it.





  • Issues of rotation 

    I received a phone call from another garage: 'We've seen you in the Top Technician magazine and are wondering if you would be interested in looking at an ABS fault for us?' The call went along the usual lines, can the symptoms be recreated? What is the repair history? The vehicle was booked in for me to take a look.

    The car in question was a 2011 Honda
    CR-V, which had been taken as a trade in at a local garage, the fault only occurred after around 50-70 miles of driving, at which point the dash lights up with various warning lights. The vehicle had been prepped and sold to its new owner unaware a fault was present.

    Fault-finding
    After only a few days the fault occurred and the vehicle returned to the garage. They had scan checked the vehicle and the fault code ‘14-1- Left Front Wheel Speed Sensor Failure’ was retrieved. On their visual inspection, it was obvious a new ABS sensor had already been fitted to the N/S/F and clearly not fixed the fault. Was this the reason the vehicle had been traded in? They fitted another ABS sensor to the N/S/F and an extended road test was carried out. The fault reoccurred. This is when I received the phone call; the garage was now suspecting a control unit fault.
        
    My first job was to carry out a visual inspection for anything that was obviously wrong and had possibly been over looked: correct tyre sizes, tyre pressures, tyre tread and excessive wheel bearing play. All appeared ok. The ABS sensors fitted to this vehicle are termed 'Active' meaning they have integrated electronic and are supplied with a voltage from the ABS control unit to operate. The pulse wheel is integrated into the wheel bearing, which on this vehicle makes it not possible to carry out a visual inspection without stripping the hub.

    Endurance testing
    With the vehicle scan checked, all codes recorded and cleared, it was time for the road test. Viewing the live data from all the sensors, they were showing the correct wheel speed readings with no error visible on the N/S/F. The road test was always going to be a long one, fortunately at around 30 miles, the dash lit up with the ABS light and lights for other associated systems; the fault had occurred. On returning to the workshop, the vehicle was rescanned, fault code '14-4 - Left Front Wheel Speed Sensor Failure’ was again present. Again using the live data the sensor was still showing the wheel speed the same as the other three, so whatever was causing the fault was either occurring intermittently or there was not enough detail in the scan tool live data graph display to see the fault. It was time to test the wiring and the sensor output signal for any clues.
        
    Using the oscilloscope, the voltage supply and the ground wire were tested and were good at the time of test. I connected the test lead to the power supply wire and using the AC voltage set to 1V revealed the sensors square wave signal. Then rotating the wheel by hand and comparing the sensors output to one of the other ABS Sensors, again all appeared to be fine. A closer look at the signal was required, zooming in on the signal capture to reveal more detail; it became easier to see something was not quite right with the signal generated by the sensor when the wheel was rotated. With the voltage of the signal remaining constant, a good earth wire and the wheel rotated at a constant speed the signal width became smaller, effectively reporting a faster speed at that instant, not consistent with the actual rotational speed of the wheel. It was difficult to see the error, zooming out of the capture to show more time across the screen it could be seen that this appeared in the signal at regular intervals, although not visible all the time because it was such a slight difference. Using the cursors to measure between the irregular output and counting the oscillations, it was clear that it occurred at exactly the same interval every time. It had to be a physical fault on the pulse wheel.
        
    This meant a new wheel bearing was required. The vehicle was returned to the garage as they wanted to complete the repair, a new wheel bearing was fitted and extended road testing confirmed the vehicle was now fixed.

  • Under pressure 

    Even apparently simple problems require thorough investigation if you want to diagnose faults right the first time

  • Your Car – Your Choice awareness widens 

    The IAAF’s Your Car - Your Choice pilot project in Chesterfield has taken another step forward following the first ever garage awareness day.More than 100 people attended the event, held at Bilstein Group, Markham Vale as garages were able to find out more about their rights under Block Exemption Regulations (BER) and hear presentations from automotive aftermarket suppliers supporting their right to service and repair vehicles within the warranty period.

  • IAAF welcomes FIGIEFA ExVe ‘Proof of Concept’ testing  

    The IAAF has welcomed the news that FIGIEFA, together with other stakeholders, is conducting ‘Proof of Concept’ testing of a ‘Remote Diagnostic’ use case as part of the investigation of the Extended Vehicle (ExVe) concept.


Search

Sign Up

For the latest news and updates from Aftermarket Magazine.


Poll

Where should the next Automechanika show be held?



Facebook


©DFA Media 1999-2019