Impact of Windshield Optical Aberrations on Visual Range Camera Based Classification Tasks Performed by CNNs

Cameras operating in the visual range of the electromagnetic spectrum are central to advanced driver assistance systems (ADAS). Front cameras, analyzing traffic, are often located behind the windshield to detect and classify objects.Thus, the area of the windshield within the camera’s field of view is a part of the optical system. Simple windshields consist of two curved glass surfaces connected by a thermoplastic interlayer. Due to defects present in the raw glass, as well as those introduced during the bending and lamination process, windshields will have optical aberrations. While optical quality may be suitable for human vision, it can fall short of what is needed for machine vision. In this article we investigate how the optical aberrations generated by laminated safety glass (LSG) influence the optical performance of a camera system and based on this, how the classification of image content by a convolutional neural network (CNN) is affected. A method for wavefront measurements of LSG samples is presented, which allows us to parameterize a linear optical model in Zernike Space. From this, we derive space-variant point spread functions (PSFs) and apply those to the dataset to simulate the windshield’s impact on the camera image. As a use case, a CNN was trained on the unmodified dataset and compared to the modified versions with the LSG models applied. We measured and modelled two different LSG samples, one with high and the other one with low optical quality. We compare the prediction accuracy of the classification with the unmodified data. The highquality sample had negligible effect on the overall classification accuracy, while the low-quality sample lowered the prediction accuracy by up to ten percentage points due to the optical aberrations.

The Recycling of Waste Laminated Glass through Decomposition Technologies

Laminated glass is ever more frequently used nowadays. This applies to the automobile industry and the construction industry. In cars, this refers mostly to the front and rear windows, whereas in construction, technical safety glass is used for railings and window glass. The task of this type of glass is to provide sufficient resistance against mechanical impact and unpleasant weather conditions. At the same time, if it is damaged, it has to break into the smallest possible pieces, or, wherever possible, the glass should remain intact on the interlayer film to prevent shards from injuring people and animals in the immediate vicinity. The paper deals with the recycling of laminated glass, especially with the effective separation of glass (in the form of cullet) from the polyvinyl butyral (PVB) interlayer film. The experimental research is focused on the mechanical separation of glass from the interlayer film by vibration, and also on the chemical cleaning of PVB film in order to allow subsequent recycling of both materials. The results quantify the efficiency of mechanical separation in the form of weight loss of the sample of laminated glass and define the particle size distribution of glass cullet, which is an important parameter in the possibility of glass recycling. The research leads to a methodology proposal for the separation of glass and PVB film and the design of equipment for this method.

Coatings with recycled polyvinyl butyral on polyester and polyamide mono- and multifilament yarns

Polyvinyl butyral is used in safety glass interlayers, mainly in car windshields. Legislative regulations require a recycling of cars after their lifetime and therefore also their safety glass. This causes the availability of recycled polyvinyl butyrate (r-PVB) originated from safety glass interlayers. Due to deteriorated optical properties, such as the transparency, and unknown amounts of plasticizers, it is challenging to reuse the recycled material in new windshields. Therefore, it is of particular interest to find new fields of application for r-PVB, such as the usage as a textile coating. In this research, r-PVB was investigated as a material for yarn coating. Polyester and polyamide mono- and multifilament yarns were coated continuously with solely a polymer dispersion and with mixtures of crosslinking agent and polymer dispersion. Crosslinked r-PVB coatings showed enhanced properties toward abrasion and chemical resistance. Coatings without the crosslinking agent showed a diminished abrasion resistance and could be washed off with ethanol. Mechanical properties of the monofilaments were influenced by the r-PVB coating in general. However, varying concentrations of the crosslinking agent did not affect the mechanical properties.

Glass-paper-laminates: examination of manufacturing methods, properties and discussion of potentials

This paper examines various ways of combining paper and glass as a laminate and the effects on transparency. Laminate in this context means a layer of paper sandwiched in between layers of glass held together with an adhesive. Different kinds of papers and adhesives were used to study the potentials of glass-paper-laminates as related to transparency and translucency. These laminates may find applications in building construction, safety glass, printed electronics and more. Utilizing a variety of adhesives and papers, qualitative evidence found that epoxy achieves the best with regards to transparency and adheres most effectively.

Collision mechanism of unmanned aerial vehicles onto glass panels

The international standards for safety glass design have been limited to the use of a soft body impactor representing a human body to estimate the accidental collision resistance. However, UAVs have recently emerged as a new collision source due to their increasingly widespread use. An evaluation protocol for UAV collision is required to secure the safety of the glass used for structural cladding. In this paper, the collision mechanism of unmanned aerial vehicles (UAVs) onto glass panels is investigated by conducting experiments. We evaluate the impact forces of UAV collision on the glass panels and verify the numerical simulations by comparing the resultant forces from the experiments. Using the verified numerical simulation, 72 different collision conditions (changing the impact angles and velocities, respectively) are modeled, including extreme collision conditions that could be infeasible in the real experiments under manual control. Through this study, complicated collision mechanisms are clarified and elucidated. Additionally, specific collision conditions that induce maximum and minimum threats are identified, which should assist the design of future structural cladding glass systems and the UAV operation discipline in urban areas.

Batch Assignment of Parallel Machines in an Automotive Safety Glass Manufacturing Facility

In this paper, optimization algorithms are used to solve a batch assignment problem of parallel processing furnaces in American Glass Products (AGP), a world leader company in the design and manufacturing of curved armored glass for transportation purposes. The problem consists in optimizing the bending process, which is considered to be the bottleneck workstation in the armored glasses production line in AGP. The objective is to maximize the efficiency of the furnaces and minimize the tardiness delivery of orders. Due to the complexity and constraints of the problem, we developed a proper dispatch algorithm and a Tabu search technique. The results are encouraging: the indicators of furnace usage hours and tardiness delivery improved by 32 % and 7 %, respectively compared to the decisions made in the plant during an actual production week. This work was the winner of an operation research challenge between around 100 graduate students. The challenge was organized by Javeriana University and AGP.

Investigations on the execution and evaluation of the Pummel test for polyvinyl butyral based interlayers

Laminated safety glass (LSG) is increasingly used as structural element in buildings. Of central importance for safety are the adhesion and the residual load-bearing capacity in the post fractured state. In literature a large number of tests to assess adhesion is mentioned. These include, e.g. peel tests, through-cracked-tensile/-bending tests, VW-pull tests and compressive shear tests. However, especially in industry, the Pummel test is widespread for determining the quality of adhesion in LSG with polyvinyl butyral based interlayers. This test method proves to be simple and quick to carry out: The laminate is stored at − 18 °C and then completely destroyed at room temperature with hammer blows. The adhesion level (0–10) is determined by visually comparing the adhering glass fragments with reference pictures or with the help of diagrams and tables which indicate the Pummel value as a function of the free film surface. Pummel value 0 is to be interpreted as no adhesion and Pummel value 10 as very high adhesion. Due to the lack of standardization, the execution and evaluation is very much dependent on the test institution and executive person. This paper shows different Pummel classifications that can currently be found on the market. Subsequently, approaches to the automatization and standardization of the execution and especially the evaluation of the Pummel test are shown. Three image evaluation methods in Matlab are presented, discussed and compared: (1) analysis of binary images, (2) statistical evaluation of the greyscale images and (3) texture analysis using co-occurrence matrices.