Introduction:

Background

The current method that addresses the problem using image processing alternatives includes components such as HOG, LBP, and WT. The designed methodology separates the bikes from the images by the most probable area where estimation plants are used Helmet will be present, and then feed it to the abstraction and matching machine function. Chiverton et al., (2020) suggests that using a circular arc to identify the helmet in a video feed it has very low precision. Conversely, given the number of Vehicles at the speed of a given instance, the calculation which It is very demanding and needs a lot of energy. This method specifies any circular object around the bike riding a helmet.

Justification

The motorcyclists have to wear helmet to ensure a legal trip (“Bicycle road rules and safety”, 1993). People generally don’t wear helmets for a few days now since they don’t understand that fact and our plan address this issue by making all the riders wear helmet. Whenever a rider has to ride a bike, he / she may have to wear the helmet, then only the bike will start to activate. Our process also helps government and police officers to take action towards those who violate the law and penalize those who don’t wear helmet and force the riders to wear helmet. While the government’s rules require the riders to wear helmet, it isn’t worn by the riders. My research data would be helpful for them. It is been argued that coal miner helmet which is similar to smart helmet reflects on diverse sensors which will have radiations (Bhutto, Daudpoto, & Jiskani, 2016). This research paper is saving a lot of lives from head injuries. The goal of this research report is to make the riders wear helmet without using sensors and thus make it free from radiation.

Gap Analysis

An advanced neuron system based automatic visual system is expected to examine the camera view of a motorcycle, spot the registered plate and recognize the registered code of the motorcycle (Draghici, 1997). This will help to capture all on road camera to detect all the law-breaking incidents. The algorithm proposed by Sorin draghici et al., (1997) has three key components:
Plate-area extraction
Character breakdown
Recognition of characters on the plate.

Fringe the analytics component uses detection methods and smear analytics to extract the plate region, smear algorithm, anatomical and sorting methods are also finally used to recognize plate characters using statistical based template matching. The execution of the device algorithm proposed was tested on the real pictures (Chakraborty & Parekh, 2015).

Research Question:

There are the following research questions:

How to show a research meets those formal characteristics it wants?
How to ensure that a research that complies with its formal specifications has no unnecessary behaviours and consequences?
How to get rid of sensors to make sure helmet is radiation free?
How to make effective human oversight of an AI system possible after it begins to function?

Methodology:

Literature Review

Certainly, the main aspect of the research was to find how AI can detect wearable helmet on the head of the rider without sensor radiations. To study that heaps of journal articles were searched through the key and peripheral tools. Key tools used to do the research were CSU primo search library, Journals and google scholar. On the other hand, keywords with the combination of logical operators was helpful to narrow down the research paper which was shortlisted from IEEE papers. Keywords with combination of logical operator were like for e.g. “Smart helmet with AI”, “AI in transportation” AND “Bikers using smart helmet” OR “Safe riding with AI helmets”. CRAAP test was used to check the reliability on the resources to make sure correct path towards answering research questions. 13 articles were shortlisted to make sure they answer all the research questions and according to that journal synopsis was created.
Amongst all 13 papers one with titled “Machine Vision Techniques for Motorcycle Security Helmet Detection” states that video clips were taken from the roadside and transformed into photographs and it first discovers how many members are on the two wheelers and It then determines whether or not the rider wears a helmet using the KNN (K Nearest Neighbour) classifier based on circularity characteristics, average shades, average intensity of each head quadrant. (“Machine vision techniques for motorcycle safety helmet detection – IEEE Conference Publication”, 2014).

Search Strategies and Data Collection

After examining of many research papers and journal articles, Paper titled “Exploring public wearable display of wellness tracker data” helped me in developing the knowledge about how we could detect the presence of helmet on rider’s head and also maintaining track on his/her wellness with AI aspects (Colley, Pfleging, Alt & Häkkilä, 2020). The automated helmet detection strategy is as seen in Figure below.

(Marayatr & Kumhom, 2014)

Moreover, the other shortlisted research papers also state some of the key featuring elements which are needed to archive the aim. Multiple next generation technologies are required such as Head-Up-Display (HUD), Artificial Intelligence (AI), and the Internet of Things (IoT) to broaden their variety of products, consumer base, and geographic reach which will make bikers interaction easier in remote areas (Martinez-Hernandez, Boorman & Prescott, 2017).
Corporations like Sena Technologies Inc.; Fusar Technologies Inc.; and Daqri LLC have acquired significant fair stocks because to their dynamic and effective research and development efforts. Such corporations generally emphasis on innovative products and developing a worldwide possibly the best-established partnership in numerous nations such as the U.S., India, China and Brazil. (Markets, 2020)

Data Analysis

In 2018, The worldwide market for smart helmets was estimated at USD 372,4 million and is predictable to enlarge from 2019 to 2025 at a CAGR of 18.6 per cent. Such helmets are currently gaining prominence among bike riders and people on bikes for improved health, protection and ease. It is commonly used in numerous sporting events, military and fire-fighting operations. The demand is expected to drive considerations such as tight rules on road safety established by governments, increased demand for advanced wearable technology and knowledge of individual hygiene.

(“Motorcycle Helmet Market Size, Share | Global Industry Report, 2025”, 2020)

Result
By developing such technique and using the AI functionality in this crucial sector will not only decrease the number of accidents but, also will create an awareness amongst all the riders. This will also help in boosting the helmet production and its users. After analysing the plenty of articles, credibility and quality was checked by finding whether this article is from credible sources, Does the author’s tone make sense. Moreover, these are the articles useful for current research.
Now, there can be various possibilities about the usage of the helmet, using with the sensors or with the sole cameras. Here, both are advisable as they are tested in the lab before going on actual product release and tested thoroughly. Even if the rider forgets his helmet and want to ride in an emergency situation, he can’t ride due to its unique feature of scan-n-start. To solve the problem, we can propose one more solution as the rider can hang or attach the helmet on to the bike using a safety lock and can be accessible only by his biometrics recognition (Ghasemzadeh, Fallahzadeh & Jafari, 2016).

Discussion

In this section, the four research questions of my report have been answered in the detailed description, which is as follows,
How to show a research meets those formal characteristics it wants?
A proper formal research which is based on the literature and peer review usually comprises of following objectives:
Systematic: These are the steps which can lead to a predictable outcome if the reader gets familiar with the topic. This is also achieved by providing a prior knowledge to the reader using Introduction or Abstract. The flow of the research determines its easiness and predictability which is judged by the reader on the basis of ease of reading and information retrieval.

Organizational: The formal research must follow the standard guidelines comprising of following points in sequential manner,

Abstract
Introduction
Methodology
Literature review
Result and Discussion
These few points which are included in the research report for the literature review can cover the overall characteristics which an ideal review is build.

Questions:
The literature review and methodology are the main components of the research which gives a broad idea of what the author actually want to put forward any solution for the given situation and questions asked.

Finding answers:
This is the main outcome of the research of finding answers to the questions provided which can solve the issues effectively.
Here, in my research, I have covered all the necessary points which needs to be formally completed in a research topic and comprises of various benefits using the idea. This research will be proved beneficial in numerous ways which, furthermore, incorporates ideas and examples using the AI functionalities.

How to ensure that a research that complies with its formal specifications has no unnecessary behaviours and consequences?

This system gives an idea of how many traffic offenders there are in one area. It generates a database of all bike riders driving without having to wear a helmet along with a proof snapshot. Utilizing independent and accessible technology such as tensor flow, OpenCV, and tesseract makes the software relatively cheaper. This device was checked under equal lighting conditions to provide full proof and accurate performance. Such wearable helmets are integrated with AI for ageing population and also encourages people to use it for safety purpose. As ageing population may turn out as a workforce risk (Lavallière, Burstein, Arezes & Coughlin, 2016). To sum up with, Public understanding of the overall effect of the program will increase.

How to get rid of sensors to make sure helmet is radiation free?
The paper titled “EEG signal-processing framework to obtain high- quality brain waves from an off-the-shelf wearable EEG device” helped me in solving this problem. The article proves the point that wearable helmet is able to carry EEG operations without any radiations. This paper also makes sure that it monitors phycological status of the rider. This article makes helmet radiation free and gather all high-quality information which are mandatory for rider’s safety. To maintain this high-performance EEG devices powerful hardware components are required to soften real-time restraints. The article tittle “Software and Hardware Requirements and Trade-Offs in Operating Systems for Wearables: A Tool to Improve Devices’ Performance” shows they could be maintained using additional hardware components for maintenance purpose (J. P. Amorim, C. Silva & A. R. Oliveira, 2019). 
 On the other hand, for the rider’s helmet cloth will be embedded with AI aspects to gather heart rate, temperature and stress level through the smartphone (Fernández-Caramés & Fraga-Lamas, 2018). It is also been noticed that AI based wearables provides data-driven solutions for safety. Sweden decided all blue collared employees should be provided with safety as there work involves tricky things. Applying the same logic to the biker’s helmet would help to configure data driven helmet that sensor radiation helmet.

(Atiqur & Khan, 2020).
How to make effective human oversight of an AI system possible after it begins to function?
With the introduction of an AI system in the helmet to supress bikers head injury during accidents will be one kind of assistance to them. This assistance will be there with them forever until they stop using the AI built helmets. As wearable sensing devices are already in use with the construction industries which demonstrates the best use of helmets to protect and safe work environment for construction workers. It also reduces the accidents happening around the work sites (Nnaji, Okpala & Awolusi, 2020). Now, regarding the human oversights of an AI in this particular topic can be managed by some particular properties mentioned in the methodology. As, basic human nature is to forget the common things in the life no matter how crucial they are to one’s life due to the supplement of some additional features to comply with the comfort zone. This builds up a human oversight regarding that thing which can happen in this case as well. Here, I have provided some solutions to this problem. In first place the rider will not be able to start his bike without a helmet. Secondly, if the rider removes his helmet while riding then the bike will automatically raise an alarm and gives rider a warning of turning off the fuel supply so that the bike will slowly come to rest. This will safely stop the bike to function and will not have any immediate effect on the ride as the bike will not stop immediately. Moreover, helmet lock is also provided with the bike so as the rider will not forget the helmet and can be wear at the same point.

 

References

Bhuttoa, G. M., Daudpotoa, J., & Jiskanib, I. M. (2016). Development of a wearable safety device for coal miners. International Journal, Volume 7, No. 4

Bi, C., Huang, J., Xing, G., Jiang, L., Liu, X., & Chen, M. (2019). Safewatch: A wearable hand motion tracking system for improving driving safety. ACM Transactions on Cyber-Physical Systems, 4(1), 1-21.

Bicycle road rules and safety. (1993). Retrieved 28 May 1993, from https://www.qld.gov.au/transport/safety/rules/wheeled-devices/bicycle#helmets

Chan, M., EstèVe, D., Fourniols, J. Y., Escriba, C., & Campo, E. (2012). Smart wearable systems: Current status and future challenges. Artificial intelligence in medicine, 56(3), 137-156.

Chiverton, J. (2012). Helmet presence classification with motorcycle detection and tracking. IET Intelligent Transport Systems, 6(3), 259.

Colley, A., Pfleging, B., Alt, F., & Häkkilä, J. (2020). Exploring public wearable display of wellness tracker data. International Journal of Human-Computer Studies, 138, 102408.

Dunn, J., Runge, R., & Snyder, M. (2018). Wearables and the medical revolution. Personalized medicine, 15(5), 429-448.

Draghici, S. (1997). A Neural Network Based Artificial Vision System for Licence Plate Recognition. International Journal of Neural Systems, 08(01), 113-126.

Fernández-Caramés, T. M., & Fraga-Lamas, P. (2018). Towards the Internet of smart clothing: A review on IoT wearables and garments for creating intelligent connected e-textiles. Electronics, 7(12), 405.

Ghasemzadeh, H., Fallahzadeh, R., & Jafari, R. (2016). A hardware-assisted energy-efficient processing model for activity recognition using wearables. ACM Transactions on Design Automation of Electronic Systems (TODAES), 21(4), 1-27.

Jebelli, H., Hwang, S., & Lee, S. (2018). EEG signal-processing framework to obtain high-quality brain waves from an off-the-shelf wearable EEG device. Journal of Computing in Civil Engineering, 32(1), 04017070.

JP Amorim, V., C Silva, M., & AR Oliveira, R. (2019). Software and Hardware Requirements and Trade-Offs in Operating Systems for Wearables: A Tool to Improve Devices’ Performance. Sensors, 19(8), 1904.
Khan, R., & Atiqur, M. (2018). Feasibility Analysis of AI based Wearable Data-driven Solution for Safety and Health in Sweden. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-239442

Lavallière, M., Burstein, A. A., Arezes, P., & Coughlin, J. F. (2016). Tackling the challenges of an aging workforce with the use of wearable technologies and the quantified-self. Dyna, 83(197), 38-43.

Martinez-Hernandez, U., Boorman, L. W., & Prescott, T. J. (2017). Multisensory wearable interface for immersion and telepresence in robotics. IEEE Sensors Journal, 17(8), 2534-2541.

Nnaji, C., Okpala, I., & Awolusi, I. (2020). Wearable Sensing Devices: Potential Impact & Current Use for Incident Prevention. Professional Safety, 65(04), 16-24.

R. Waranusast, N. Bundon, V. Timtong, C. Tangnoi and P. Pattanathaburt, “Machine vision techniques for motorcycle safety helmet detection,” 2013 28th International Conference on Image and Vision Computing New Zealand (IVCNZ 2013), Wellington, 2013, pp. 35-40

Smart Helmets Market Size Worth $1.2 Billion By 2025: CAGR: 18.6%. (n.d.). Retrieved May 21, 2020, from https://www.grandviewresearch.com/press-release/global-smart-helmets-market
Verma, P. (2018, July 24). IIT-Hyderabad develops Artificial Intelligence to catch bikers without helmets. Retrieved May 21, 2020, from https://economictimes.indiatimes.com/tech/software/iit-hyderabad-develops-artificial-intelligence-to-catch-bikers-without-helmets/articleshow/65109382.cms?from=mdr