Spring 2019
1: Computer Vision Parking
In 2017, over 45,000 students were enrolled at the University of Houston. The vast majority of which are commuters, with about 85 percent of students driving to class as reported by the University of Houston’s Student Commuter Service. Parking has become very strained at peak hours creating additional stress and wasted time for students. The current resource for students and faculty is only available in the garages around campus. Where it displays the number of spaces available on a specific floor or they have red and green lights to indicate whether or not a space is occupied by a vehicle. There are no available resources for students to find empty spots in the currently zoned lots since the availability map that used to be available has been removed, other than having a parking pass to a specific zone. We will be utilizing advances in computer vision and open-source resources, that would integrate software that analyzes the cameras at a parking lot and identifies empty spaces. With this project, we will be implementing our software into already existing camera systems for outdoor parking lots. Which will help reduce the amount of stress and time wasted by students when searching for an open parking space.
2: TCAS Communication Encryption
Every day, about 28,000 flights are handled by air traffic controllers and “at any given moment, roughly 5,000 planes are in the skies above the United States [1]” which creates a lot of traffic in the air. There are multiple airline companies that contribute to travel needs to shuttle people across countries and continents. However, none of these companies have any security measures for communication. Since the Überlingen mid-air crash in Germany circa 2002, the TCAS (Traffic Collision Avoidance System) has been adopted by virtually all airline companies to prevent a tragedy like that from happening again. The TCAS program has no protection or authentication system for communication which can lead to a malicious attack that may cause another mid-air collision resulting in catastrophe. We plan to create a preventative encryption method that will implement a one-way authentication system for the TCAS program to allow planes to securely communicate with each other to ensure the safety of everyone on board.
3: HELI-CAPTURE GUARDIAN
In the US, 723 children ages 12 years and under die in motor vehicle accidents yearly. More than 618, 000 children ages 0-12 are not using a safety belt, car seats, or booster seats inside a vehicle. Out of the 618, 000 children that die, about 35% were not using their seatbelt. Another safety concern is children who die from hyperthermia while trapped inside an unattended vehicle. There are 38 children on average that die yearly from hyperthermia when left inside a vehicle and about 900 have died in the past 10 years.
The safety of children is one of the most important values that our society holds. There are many different ways in which we attempt to keep a child away from any danger. A problem we face in a child’s – is safety within a vehicle. When children are passengers they may not be buckled in or the parents will forget their child in the vehicle. If a child is not wearing a seatbelt, it reduces their safety in a car collision. When a child is left unattended in a vehicle, they face the danger of possible death or injury due to a heat stroke. The team wants to reduce these scenarios with a system that detects whether the child is wearing a seatbelt or not or if it is properly worn with the implementation of the Child Monitoring Safety System. While the belt is attached, the system will calculate if the belt is putting any strain or danger on the child’s neck by using object and facial recognition. This system will also detect and notify whether the child was forgotten in the vehicle. Furthermore, the system will help drivers to be less distracted with a discontented child in the backseat. Facial recognition will be used to read the child’s facial expression and display the status of the child for the driver to see without having to turn around and taking their eyes off the road.
4: AUTOMATED DIRECTIONAL MODULE TEST INTERFACE
QES Directional Drilling, a company focusing on providing a wide range of completion, production, and drilling services reached out to the team to propose a project that would solve their technician’s time spent on calibrating their measuring while drilling (MWD) modules. If it is possible, reduce the time taken to calibrate their directional module by at least 50 %. Currently, the directional module is being manually rotated to achieve desired orientation readings and it is performed by a senior-level technician that takes from 20-30 minutes to test each module. This process is tedious due to the constant monitoring and rotating of the tool. The tool must be manually rotated to the desired angle and it must land on the desired clearance of that angle. When it lands within that window, the tool must be stable for a certain amount of time and if interrupted, the test must be restarted. This procedure tends to leave room for human error as the current clearance for technicians has been set to 3 degrees; the team plans to minimize that clearance to 1 degree. The team proposed an automated calibration routine that uses LabVIEW, a stepper motor, and a DAQ. Providing QES with a new automated calibration routine will help the normal timing period it takes for each module to be tested to be reduced by more than 50% with high accuracy, no interruptions, and minimal supervision from the senior-level technician. This method will lead to an increase in customer satisfaction, faster testing and delivery time.
5: Smart Incubator
Plants are one of many sources used to sustain life. From those plants, we are able to grow natural food products like vegetables. Within the plant species, there is one particular type of plant that although it might be difficult to grow due to weather and climate change it contains a vast majority of protein, minerals, and vitamins which can be beneficial towards certain research labs and medical uses. The team will be trying to grow mushrooms (fungi), specifically Oysters in a specific environment controlled area called a Smart Incubator. The Smart Incubator can be used to grow several types of mushrooms with its several sensors and user-friendly interface. In this case, Oyster mushrooms need certain limits to grow within an environment. Specific levels of humidity, temperature, light, and carbon dioxide can be used and altered within the Smart Incubator to achieve greatness.
6: CYHM – Can You Hear Me
The inspiration for our project came from Mark’s younger brothers who were both diagnosed with a developmental disorder known as an autism spectrum disorder. This disorder targeted their communication abilities which led to the idea of making communication between verbal and nonverbal people much easier. After our interview with the Life-Long Learning Center, we learned about the potential impact our project can have on the scale of personalizing communication types for not just ASD individuals, but also for nonverbal communication as a whole.
After conducting more research, we found that 55 percent of communication is actually performed through gestures. Our project will attempt to narrow the communication gap of non-verbal patients diagnosed with autism as well as individuals with ASD; whether it be from a symptom such as ASD or from other factors such as autism or deafness, we will create an application that translates gestures and actions into words and voices.
7: IMPLEMENTING FIBER OPTIC IN THE FG8800 SYSTEM
Upstream oil and gas companies require fire control panels in order to detect and mitigate fire hazards in Class I, Division 2 environments. This classification indicates the presence of volatile flammable liquids or flammable gases is being handled or processed. When such an environment exists, monitoring systems send information to the control panel, which dictates what steps need to be taken to control the situation. In the oil and gas industry, these systems would be sending information from platforms hundreds of kilometers away, which can cause problems if the control panel is using inadequate hardware. According to Allestec, 25% of their oil and gas customers need a fiber optic option to connect to systems in a remote area to provide better communication.
Currently, Allestec fire panels are equipped to communicate over a Controlled Area Network Protocol (CAN bus). This kind of protocol is ultimately designed to allow microcontrollers or other devices to communicate with each other within applications without the need for a host computer. Using CAN bus, Allestec can reach a maximum effective distance to a node at 1000 meters at a speed of 50 kbps. This is a great inconvenience in the oil and gas industry because offshore oil platforms can be as far as 200 km or greater from the coast. Wireless communication is not plausible due to the high interference caused by metalwork on the platforms. The best alternative would be using fiber optic communication because by design it can transmit data over great distances before attenuating.
8: Smart Airflow System
Heating, Ventilation, and Air Conditioning (HVAC) systems are one of the largest consumers of electricity in buildings, accounting for about 40% of the total energy consumption in buildings. In order to operate, HVAC systems may require the use of fans, cooling devices, heating devices, pumps, and cooling towers. Of the total energy consumed by the average HVAC system, 34% of that energy goes toward fans, 27% goes towards cooling, 17% goes towards heating, 16% is used by pumps, and 6% is used by cooling towers. While building characteristics such as design and location can help reduce the amount of energy consumed, a change in the way HVAC systems operate would have a more significant impact on the consumption of energy. To make a more energy-efficient HVAC system, a reduction in the amount of electricity used by each HVAC component is needed without sacrificing the comfort of the individuals inside the building.
Some implementations for controlling HVAC systems have been ON/OFF controls and Intelligent ON/OFF controls; however, those solutions come with their own drawbacks and challenges. The ON/OFF controls work by allowing users to set upper and lower temperature limits. The problem with this is that having simple ON/OFF controls can create dead zones where the HVAC system does not do anything to maintain the temperature while it is in between the lower and upper limits. The Intelligent ON/OFF controls do in fact reduce the amount of electricity used by predicting how long the HVAC system will take to reach the upper and lower limits. However, the HVAC system has a hard time predicting how long the temperature takes to reach the set limits especially if there are disturbances present in the HVAC system resulting in less efficiency.
Our objective is to create a reactive system that can modulate airflow into a room. The system would be able to recognize users, their preferences, and their patterns and use this information to control air dampers in order to manage airflow. Therefore, the system can manage temperature while reducing the amount of power consumed by the HVAC system.
9: EASY RECYCLING
In Denmark and neighboring countries, there are dedicated recycling machines for plastic bottles, which have greatly improved the percentage of plastic recycled overall. To encourage people to recycle these items, the system rewards the user with a small cash payment for each bottle recycled. However, in Denmark, every bottle sold includes an added bottle deposit. Once the bottle is returned, the tax is repaid to the consumer as the incentive. Also, every bottle manufacturer is required to include a special label, called a “Pant” (the Danish word for deposit), that corresponds to the type of plastic used and the cash value of the bottle. This system was implemented in Denmark in 2000. In 2016, about nine out of every ten bottles sold in the area were returned through these machines upon use1. There are now 3,000 of these machines in Denmark alone. Also, as of 2016, Denmark ranked 11th in the world for the percentage of waste recycled at 44% (compared to the U.S. in 18th place at 35%).
Once it is evident that the Easy Recycling system is effective at the local level, U.S. legislators may be more inclined to create legislation that implements similar bottle taxes. Perhaps, in the future, there will even be a similar system to the “Pant” labeling system. As of today, the University of Houston has 90 “Big Belly” waste and recycling receptacles that have helped to reduce waste and increase efficiency in the collection. Our system will improve on their design by attracting new users and increasing subsequent usage.
The objective of the project is to encourage students to increase their recycling habits at the University of Houston by using a system that will reward them at little cost to the University. This will, in turn, reduce the carbon footprint and set a trend for other institutions, eventually leading the city to implement it as a standard.
10: Flood Sensor Network System
Flooding has been a major problem for the city of Houston, several bayous run through the city and a lot of the city is on a floodplain. Hurricanes or heavy rainfall can cause major flooding around the city and makes it extremely dangerous to travel. During Hurricane Harvey in 2017, an estimated 39,000 people were forced to evacuate their homes due to imminent flooding; with water reaching as high as 50 inches. In their attempt to flee danger, they were faced with a new danger, flooded roads. Reports stated there were over 500,00 disabled vehicles on the road and about 100,000 stranded Houstonians on the roads. People evacuating had no idea which roads were safe and which were flooded, many of the members of our group faced this problem of not knowing how to get out safely. This is what led us to our idea of a network of water level detectors that would provide real-time data of road conditions. Our end goal is to be able to send the data from our sensors to GPS services, such as Google Maps or
Waze. The data from our sensors could allow the software to know which roads are flooded and divert users to a safer option.
11: Self-Driving Autonomous Car Project (SDACP-OBJ)
This team’s project is the Self-Driving Autonomous Car Project (SDACP-OBJ) using object detection and Semantic Scene Understanding (SSU). According to the National Highway and Safety Administration (NHTSA), 94% of vehicle crashes in the US are caused by human error and distracted driving itself accounts for approximately 25% of all motor vehicle crash fatalities.
Our goal for this project is to reduce this number significantly by automating this process through complex software algorithms that allow automotive manufacturers to offer this advanced safety technology to the customers at no cost to them. Some examples of companies using this technology are Toyota with their Safety Sense (TSS), which offers lane avoidance and object avoidance baked in all of their upcoming vehicle lineups. Over the course of this semester, our team has continued to do exhaustive research on neural networks, deep learning and how the two technologies would help us improve the working prototype of our Autonomous Car Project Algorithm that competes with hardware integrated solutions such as those implemented by Honda that uses sensor cameras in the windshield that feeds data into the onboard computer for image detection, however this solution is offered at additional cost to the customer which might alienate customers away from this fairly unproven technology in the consumer space. Our niche is curious customers that are interested in being a part of this technology but aren’t necessarily convinced to pay for it yet. With diverted and distracted drivers on the rise, potential car buyers are requesting further developed security frameworks that can help counteract mishaps before they occur. New innovation can limit the effect of blindsides, object detection alerts drivers when they veer out of a path and even apply the brakes when a mischance is up and coming.