HUD GUI – 3 Stages of Feedback (Fatigue)

Stage 1) Driver Physiological State: OK (no indications of significant driver fatigue)

GUI HUD DRIVER STATUS - OK

Stage 2) Driver Physiological State: Drained (Early indications of driver fatigue)

GUI HUD DRIVER STATUS - TIRED (DRAINED)

Stage 3) Driver Physiological State: Exhausted(Prolonged indications of driver fatigue)

GUI HUD DRIVER STATUS - EXHAUSTED

HUD GUI Interface, A New Aesthetic Approach

This is a progression from the HUD GUI as shown in earlier blogs. The Audio Equalizer and Car Diagnostics (in the older format) have been removed in favour of a driver-centric interface. A minimal yet forward-thinking approach has been utilized to complement the safety, and driver focused nature of this project. In the iteration of the HUD interface the emoticon on the top of the display will change depending the current physiological state of the driver. This emoticon will change when 3 scenarios arise:

Stage 1) Normal State – The Driver is calm and relaxed but the system does not indicate that the driver is suffering from considerable, and possibly dangerous fatigue. This is the default stage.

Stage 2) Tired State – The Drivers’ physiological condition indicate that he or she is tired. The emoticon will change in response to Stage 2, and the air temperature will be decreased by the system.

Stage 3) Prolonged Fatigue – The Drivers’ physiological state has not recovered from a tired state over a set period of time, and the system will notify the driver that he/she may be suffering from serious fatigue. If the drivers physiological state does not recover from Stage 3, an audio signal will be played until the driver takes a rest/ and or his/her physiological state indicates Stage 1 or 2.

HUD GUI

Project Presentation Poster (Draft 1)

This is a first draft of the poster that will used as a form of explanatory diagram for the reactive in-car system. The intention is to print this at A0 size and place it on the allocated 8′x8′ board prior to our assessment.

Project Presentation Poster (Draft 1)

Physiological Data Matrix to Indicate Impending Sleep State – A New Approach

Initially we sought to determine the driver’s impending sleep state by setting a threshold heart rate, which would indicate the driver’s state of relaxation, however testing of the process proved unsatisfactory due to the fact that the resting heart rate varies from person to person, and while, for example, 60 BPM (beats per minute) may indicate that one individual is very relaxed and nearing sleep, for another, fitter, individual, the sleeping heart rate could be much lower.

To counter-act this problem we have taken the “variability” approach across all the systems sensors, whereby the system interprets decreasing heart rate over time, combined with less frequent movement and increasing skin resistance as an indication of an increasingly relaxed state. This avoids the issue of setting threshold heart-rates, for example,  in order to determine the drivers physiological state, and allows greater access to the system, as even people with conditions such as  diabetes, which affect heart rate, should be able to utilize the system.

Physiological Data Matrix to Indicate Impending Sleep State

Heart Rate Monitor v3

For the 3rd version of the HRM a lot of changes were made to combine the results of the first and second version:

• The led and phototransistor in the steering wheel were removed, resoldered, and reinserted so that they were more solidly built and a much closer together.
• a deeper channel was cut out along the inside of the steering wheel for the wire to run along inside.
• Another HRM kit was bought from Maplins, assembled and tested.
• The sensor probes in this kit were then replaced with the ones from the steering wheel.
• The power source (4 x AA batteries)was replaced with one from the arduino board (5V+ and GND).
• The standard red LED on the kit was replaced with a super bright white one to give a stronger signal.
• A breadboard was used to make wiring easier.
• Using an LDR, glossy paper and some tape, an interface was created between the kit and the arduino.
• The whole circuit is designed in modules that can be added/removed as required and the whole interface is attached to a cardboard base.
• The switch on +5V is yet to be added.

Factors That May Influence Heart Rate

Fitness Level: Higher fitness level = lower resting heart-rate

Age: Younger people generally have faster resting heart rates that older people

Gender

Stress

Mental Illness

Medications: e.g. Diazepam (slows heart rate)

Existing Heart Problem:

e.g. Tachycardia (usually resting heart rate of 100BPM or higher)

BradyCardia (usually resting heart rate of 60BPM or less) may be disregarded in elite athletes with no other symptoms. Miguel Indurain, a Spanish cyclist and five time Tour de France winner, had a resting heart rate of 28 beats per minute, one of the lowest ever recorded in a healthy human.

Thyroid ailments – e.g hyperthyroidism (increases heart rate)

High blood pressure: increases heart rate (in diagnosed cases, blood pressure medications such as Beta Blockers will lower heart rate)

Alcohol – increases heart rate

Fujitsu’s PalmSecure takes high-speed, contact free biometric readings

Fujitsu has just announced a new palm vein authentication device, one that’s being touted as the world’s fasted, and the first that works without the user actually touching the device. Unlike past implementations of this technology, which moved at a comparative snail’s pace, PalmSecure can be used by people traveling down busy corridors, in as little as one millisecond. Despite its speed, Fujitsu insists that this bad boy performs with the same level of accuracy as its slow moving brethren. This technology has great implications for smart in-car systems that can monitor a drivers biometric data.

Heart Rate Monitor V2

Over the past weekend I rebuilt the heart rate monitor using parts from the original Maplin kit as well as some additional components. The main differences were:

• The 4 vehicle grade wires inside the steering wheel were replaced with a single 4 core telephone wire.
• The joints were reinforced with solder to provide better contact and reliability. The other end of the telephone wire was then connected to a circuit on a breadboard via two pcb terminal blocks, effectively creating a plug and socket which makes things easier in the long run.
• The Maplin HRM board was removed and replaced with an interface into the arduino via the breadboard. This was connected into PD via a similar patch as before and used to graph out the incoming light signal.

Results/Considerations

• Using the telephone wire instantly made the wheel less messy to construct and put together. The colour coding also eases keeping track of what I’m doing inside the wheel. This wire can also be made thinner again by stripping the outer wire from it and just using the 4 insulated cores, which are even thinner than the average jumper wires.
• What the Maplin kit did not reveal is that the sensor in question is very sensitive. In the PD screenshot you can see that the actual heart beat is very tiny in comparison to the overall scale. However once it is found it is very obvious on the graph and i have verified it by manually checking my own pulse in parallel. This I assume is what the potentiometer on the Maplin kit did.
• The sensor works in a range of 0-1. There is a large jump of 0.2 to 0.4 depending on light conditions between when the sensor is covered or not. The variation of a heart beat is much smaller as can be seen in the screenshot.
• The next step is to calculate a BPM from this array of data and replace the functionality of the potentiometer on the maplin board within PD somehow, either automatically or manually.
• A prototype of the Galvanic Skin Response will also be ready in the next day or 2. A start has been made but it would seem I need some larger resistors for both sensors(in the region of 100kOhms). As you can see from the screen shot I am already using several resistors in series for the HRM.

Wiring notes(for my own use)

Real-Time Driver Diagnostics Photoshop Mock-Up v1

Galvanic Skin Response (GSR) and ElectroCardioGram (ECG) Picture-in-Picture Readout

For the purposes of the final presentation, the concept of a “picture-in-picture” display, similar to a feature used in many modern TV sets for watching two channels at the same time, may be beneficial for demonstrating how the final prototype works. As the driving force of this project is to develop a ubiquitous sensor system whereby the driver is generally unaware that the reactive in-car system is at work, the readout of the Galvanic Skin Response (GSR) or the ElectroCardioGram (ECG) A.K.A Heart-rate, will not be made visible to the driver within the HUD.

However, for demonstrating to others how the system interprets physiological data and reacts accordingly, we could a present picture-in-picture box that displays how the relevant biometric date is interpreted in real time.

Honda Multimedia HUD Interface Test V1