AR-Bicycle. Smart AR Component Recognition to Support Bicycle’s Second Life

We live in a world dominated by two great rules: time and money. Two pillars to which we subjugate our ways of life are direct parents of consumerism, which has played and is still playing a key role in the long-standing issue of climate change. The problem of the environment has not only one cause, but is a mix of neglect and lack of foresight typical of human hedonism that celebrates the golden moments of the present without thinking about the consequences left for the future of the earth. The “do it yourself”, meaning the ability to get involved in small challenges, find yourself a handyman and be able to fix things would have a very positive impact on the environment as well as on our lives and our self-esteem. We live in the giant contradiction of being surrounded by technology that would allow us to have access to infinite resources and yet we do not seem able to "do-make" anything anymore. This paper involves several field of application: sport and well-being, the environment, technology, and the “do it yourself”, using as starter object the bicycle which is an object that is growing more and more even in large Italian cities such as Rome and Milan which have not always been bike friendly. Yet people use them, motivated perhaps by trying to reduce their environmental footprint in their own small way. Alternatively, motivated by economic reasons defined by the high cost and in recent times the exponential growth of gasoline in general due to private vehicle. The bicycle is a very interesting means of transport, has no costs and greatly reduces travel times (compared for example to moving on foot or by public transport) and lead to independence, which for many is the reason for preferring the private vehicle. The aim of the research presented in the paper is to offer a service, which allows us to simply carry out maintenance and repair actions anywhere and anytime. Trying to fill the gaps that the classic means of searching for information and tutorials have, such as slowness and difficulty in finding a specific problem. With the aid of advanced representation tools, such as Augmented Reality and cross-platform graphics engine such as Unity 3d we have the possibility to create a useful tool for identifying the components of complex elements. Using the Model Target Generator’s Vuforia feature, we have the chance to create a Model Target with an Advanced Guide View and then submit it as a training set to get the Advanced Model Target Database. This will allow the application to recognize the (generic) bicycle object and report hotspots that may need repair. By selecting the hotspots, the animation will appear in Augmented Reality, showing how to replace the component and possibly a link to shops or resources to retrieve the necessary pieces. The idea is to have a tutorial anywhere and always that is friendly and immediate, shortening the distances and the classic limits of the existing tutorials or guides. The aim is to foster and encourage, where possible and using AR capacities, the constant and autonomous maintenance of the products. In the future it would be interesting to increase the scalability of the product for now limited and tested only on Mountain and City bikes.