Heads Up Glasses
Team 13
TA: Christine Wu
M-Shop Instructor: Izzy Lambobard
Problem Statement
Non-commercial drivers need a way to be alerted when drifting into sleep or distraction in order to avoid car accidents.
Drowsy driving: driving while sleepy or fatigued
Distracted driving: driving while not looking at the road
Leads to slower reaction time and paying less attention while driving
Causes great risk, danger, and tragic accidents
How big is this problem?
Distracted Driving
3,142 deaths
8% of all fatal crashes
caused on annual basis.
Drowsy Driving
6,000 deaths
50,000 injuries
91,000 crashes
$109B costs to society
caused on annual basis.
Sources:
Drowsy driving data: CDC & National Highway Traffic Safety Administration
Distracted driving data: Insurance Information Institute
Users & Purchasers
Who are the users?
Non-commercial drivers
Anyone who drives frequently for long periods of time
Prone to distraction/falling asleep at the wheel
Don’t have resources for expensive technology
Ex: Students, Uber drivers
Target more accident-prone 16-24 year olds
Who are the purchasers?
Non-commercial drivers
Concerned parents
Incentivized by insurance agencies
State of the Art
There are many innovations that all under three main types of drowsiness detection measures.
However, each of these innovations are incredibly expensive, meaning that most drivers have no protection against drowsy or distracted driving.
Source: https://encyclopedia.pub/entry/20784
Vehicle-based technology
Lane monitoring technology
Detects when car is drifting out of its lane using a camera and notifies driver through sound, visual, and/or vibration alerts
Limitations
Price
Accessibility
Portability
Behavioral methods
Optalert driving glasses
Wearable glasses monitor drivers’ eye movement to detect drowsiness
Limitations
Price ($4000)
Maven Machines headset
Smart bluetooth headset measures head movements and mirror checks to monitor drivers’ alertness rates
Limitations
Not on the market
Designed for specific audience
Physiological methods
SmartCap
Measures alertness and fatigue by analyzing brain activity (EEG) and delvers digital alerts
Limitations
Not on the market for consumers
Embedded into wearable hard hats
Specifications
Based on user needs, we determined that cost, accuracy, and safety were the most important specifications and should be weighed 2x as heavily.
Prototypes
Initial Prototype
Idea: metal ball sits inside of a tube attached to the side of a pair of glasses. The metal ball would hit a sensor located at the front end of the tube when the user's head tilts forward and sound off an alarm.
Sketch of Initial Idea:
Video of Initial Prototype:
Working Prototype
After working with Prof. Truex, we learned about the tilt sensor which follows the same concept. We then began the circuitry process using the tilt sensor, an arduino, a buzzer as the alarm, a battery as the power source, and foam core to make our first working prototype for testing.
Final Prototype: Works-Like
As we prototyped, we decided to create a glasses attachment instead of glasses as it would be cheaper to produce from a business stand-point and more customizable for users.
After finalizing our works-like design, we created our prototype using Solid-Works and 3D printing. This prototype includes three parts: the box with clip attachment, a screw on lid, and a mounting plate which allows users to adjust the angle in which the sensor would be tilted at. Because everyone is a different height and may drive different cars, we used the mounting plate to make sure that it would be adjustable for our users for easy use.
As shown in the demonstration video, when the driver is no longer looking ahead in their line of vision, the tilt sensor would send the signal for the alarm to go off.
Final Prototype: Looks-Like
Due to time constraints, there are ways in which we wanted to improve and add to our prototype but didn't have the time to build out. Therefore, we also created a "looks-like" prototype which is ultimately what the goal of our product would be.
Additional Features of "Looks-Like" Prototype
Smaller, more compact device
Using smaller battery holders and arduinos
Uses an accelerometer as sensor rather than tilt sensor
Works in multiple planes, whereas tilt sensor only works with forward and backward tilt
ability to customize code - potential for less false alarms and higher accuracy of detection
As seen on the right, the "looks-like" prototype would still work just like the "works-like" prototype above.
SolidWorks Sketch: Box
SolidWorks Sketch: Lid
Testing & User Feedback
Testing Plan
Testing Accuracy of Detecting Driver’s Line of Vision
We tested how accurate the device was at detecting specified angle and activating alert
When the trigger angle was set at 20 degrees below eye line, the device accurately triggered the alarm 88% of the time (44/50).
Specification: accurate 90% of the time
Testing False Alarms for 4 different codes
Different sensor code we tested:
Instant alert
If (triggered) sound alarm
500 ms delay
If (triggered) wait 500 ms, if still (triggered) sound alarm
1 s delay
If (triggered) wait 1 s, if still (triggered) sound alarm
Loop check
If (triggered) check again every 100 ms for 500 ms, if always (triggered) sound alarm
Blind Spot Checking
100-Yard Deceleration
Blind Spot Test Data
Shows 1 second delay code to be the safest (meets our specification)
Deceleration Test Data
Shows 1 second delay code to be the safest (closest to our specification)
Testing Safety
All 4 group members tested driving with the prototype
Surveyed 17 testers on safety of the device
Rating: 4.58/5
Specification: Doesn’t inhibit ability to drive safely
Testing Ease of Use & Comfort
We surveyed 17 users on ease of use / comfort
Ease of use rating: 4.71/5
2-step setup process
Comfort rating: 4.35/5
Specification: comfortable, easy one-time installation or one-step setup process
Calculating Expected Battery Life of Product
Battery voltage: 6V or 367mAh
Arduino nano (19mAh) + Accelerometer (.35mAh) + Buzzer (20mAh)
Battery life = 367mAh / (19mAh + .35mAh + 20mAh)
Battery life = 9.33 hrs (for continuous buzzing)
Battery life = 19 hrs (without buzzing)
Future Considerations & Business Plan
Ethical Issues
Safety while driving
Did not find driving with alert unsafe or detector to be obtrusive (through testing)
Testing process
Testing in a stationary car
Controlled environment
Accuracy
This is a late-stage intervention - could lead to users becoming too reliant on device therefore engaging in unsafe drowsy driving practices
Plan to make users aware of this concern
Sustainability Approach
Minimal materials used for box design
Running on low voltage battery
Rechargeable batteries/power source in the long term
Sustainable packaging and shipping
Encourage recycling of electronic parts
Pay workers ethically
Business Model Summary
Variable Costs
Fixed Costs
Breakeven Sales Volume
Total Market Size = 37,707,958 (based on drivers who admit to texting and driving).
Next Steps
What could be improved about our prototype?
Increased comfort, add rubber edges
Clip adjustable for more glasses types
Explore even smaller parts to build a smaller device
Reflections
What did we learn in the process?
Big Picture Takeaways
Testing is important
Build fast and early
Materials are limited
Skills
Solidworks
Circuits (how to use an arduino and soldering)
3D printing
Acknowledgements
Special thanks to Prof. Vicki May, Izzy Lambobard and M-Shop Instructors, Prof. Tad Truex, Chris Vollmann and the Thayer Instrument Room, Christine and the TAs, and our classmates!