Advanced driver assistance systems (ADAS) such as parking assist have been wide spread. Although ADASs have successfully reduced driver's workload, it is pointed out that such ADASs do not take development of driver's skills into account. The present study focused on the assist system for parking in backward since it is a demanding task for novice drivers. We proposed a method to assist driver's parking operation using haptic shared control, which aims assisting driver's operation by haptics and improving driver's operation skill simultaneously. The proposed method displays torque around the steering wheel that is composed of vehicle position feedback to reduce its error from the desired position and steering angle feedback to reduce its error from the desired steering wheel angle determined by the preview driver model. Driving simulator experiments with various combinations of the feedback gains indicated that the skill improvement were achieved after intervention by the proposed assist that emphasizes the desired steering angle.

View full abstract

This paper proposes an operation assistance system that compensates for driving skills of elderly drivers and realizes good course tracing performance. This system consists of self-localization and assist torque calculator based on risk-potential that indicates the sense of danger that driver feels from road boundary. The shape of risk-potential is fixed by given road width and parameter study that uses maximum lateral acceleration as a basis. The assist torque is calculated from the lateral force that is obtained from the shape of risk-potential and the lateral displacement at the preview point. The amount of assistance will be determined by multiplying calculated assist torque and assist gain that indicates the ratio of assistance. In this paper, the assist gain is uniquely specified and the effectiveness of proposed operation assistance system is verified by conducting experiments of shared driving using the driving simulator in crosswind scenario.

View full abstract

Recently, the number of traffic accidents is decreasing due to the development of safety technologies. However, it is difficult to select an appropriate speed to pass the unsignalized intersection without a regulation stop sign for elderly person. To support such drivers, this study aims to construct an expert driver model based on risk potential field. First, driving data of expert drivers are acquired by the experiment in the public road. Next, by optimizing the parameters of risk potential function in the framework of an optimal theory, the velocity profile of the expert driver can be estimated. Finally, the feasibility of the proposed model is verified by comparing the simulation results with the actual driving data by expert drivers.

View full abstract

In this paper, we first define a friction model as an extension of the 2D-Coulomb model, considering nonholonomic constraint. Next, we formulate steady-state cornering and show that turning radius can be decreased by drifting. The feasibility is tested under simulation.

View full abstract

This paper presents a system to discriminate driver's posture using face tracking for driver support. A camera is arranged near a room mirror of a car, since the setting does not conflict with the law and is easy to implement. A driver's face region is firstly detected by Template Matching. The region is tracked using color information. Cam-Shift Tracking or Mean-Shift Tracking is used for the face tracking. Finally, driver's pasture is discriminated using the location of the face in images. The area for the discrimination is set up empirically. The effectiveness of the system is evaluated by experiments for a driver wearing glasses, a cap, or a mask.

View full abstract

There is a surgical method called Interventional Radiology (IVR). We have been developing surgery support robot of IVR. IVR has the advantage that invasiveness is low, therefore the number of IVR surgery has been increasing in recent years. In lung cancer treatments, however, it has the problem that doctors are exposed to radiation because IVR is conducted in CT equipment. The purpose of this study is to avoid the problem by operating remote steerable IVR robot. IVR robot with 5 DOF have been developed in the prior study. Previously, each joint was controlled independently. Therefore, when the posture of needle is changed, the position also should be corrected manually. To improve operability of IVR robot, the forward kinematics and the inverse kinematics is solved. Finally, accuracy of needle tip position is evaluated.

View full abstract

We have been developing a laparoscopic surgical training system. To train operations under various situations, preparation of various organ model is required. For example, shape, hardness, tactile and tumor location, size of organ. In this paper, we introduce a casting process for making a soft organ model with virtual tumors at a low cost.

View full abstract

Myoblast cell sheets are taking into the clinical treatment of heart disorders. In our research groups, we have developed a surgical robot system, includes a robot system which can cancel and compensation the beating heart surface movement for the myoblast cell sheet transplantation. The purpose of this study is to develop a transplantation mechanism with cell sheet scooping mechanism mount on our surgical robot and evaluate usability of this robot system. An experiment is conducted to evaluate the transplantation and force of our robot system in tracking a beating heart surface movement. From the experiment, we got the results that the transplant success rate is 88% and the shearing force reduced approximately 40% than the transplantation by the subject, and the pressure reduced 30%.

View full abstract

High intensity focused ultrasound (HIFU) has a potential for a treatment of stones or tumors. In HIFU therapy, respiratory organ motion makes it difficult to focus HIFU on an affected area, and increases the risk of damaging healthy tissues around the target. In this paper, we proposed the method for tracking a respiratory organ motion, in which we use a visual feedback and a prediction model of a respiratory organ motion, and realize high accurate robot position control. The prediction model is made from the most recent organ motion, and always updated. In the experiments, respiratory kidney motion of two subjects can be successfully tracked with an accuracy of 0.88 mm.

View full abstract

We have developed retractor for Ultrasound Imaging in Water Filled Laparoendscopic Surgery(WaFLES). However, it is hard to keep three dimensional space for surgical treatment in WaFLES using the retractor because it is plane. Plane exclusion can't prevent organs' flow in space in WaFLES. Therefore retractors that enable to exclude organs three dimensionally is needed in WaFLES. The purpose of this study is to develop a new retractor that can extend and keep three dimensional space for surgical treatment in WaFLES without affecting ultrasound image. Experimental results show this retractor did not affect ultrasound image, and could extend and keep three dimensional space.

View full abstract

The authors aim to detect snow on the road from sensor data for autonomous driving in winter. This paper describes collection of sensor data on snow road and trials of snow detection from LIDAR data. The autonomous navigation is affected by the snow on the road and roadsides. The place and quantity of the snow is detected by comparing 3D shapes measured by LIDAR. We collected snow data by using 3 types of LIDAR sensors (FOCUS 3D, 2 mobile platform), and confirmed that snow mountains on the road-side, which are 2 and 3 m, can be detected from the comparison. We also confirmed that snow on the road, which is several cm, can be detected from FOCUS 3D data. However, it is hard to detect them from shape data measured by mobile platform because of LIDAR measurement error and its spatial resolution.

View full abstract

Torque Vectoring Differential (TVD) is a differential device for generating a torque difference between the left and right wheels and there is an effect of improving driving performance and motion stability of the vehicle. One of author developed TVD for EV that amplifies torque difference of two motor was developed and demonstrated. In this study, we developed a control system for a small EV equipped with TVD.

View full abstract

This paper describes a prototype of car sharing system using a self-driving car. We use 3D range sensor for detect road boundaries. First, we detect curb positions and fit a line to them with RANSAC. Based on the fitting result, we classify the boundary into straight or circular, and estimate the parameter for each class. These parameters are used for controlling the car to follow the road or turn a comer. We also develop a browser-based car-calling system by which a user can call the car using map and GPS data. The car sets the user location from the specified GPS data and moves there autonomously. When the car arrived, the user switches the car to the manual-driving mode and starts driving. When arrived at the destination, he/she switches the car to the self-driving mode and just gets off. The develop system shows the feasibility of a new type of car sharing system.

View full abstract

This paper presents a set of experimental results of force perception abilities in operating driving interfaces by the limbs with the developed driving simulator. The results show that force perception characteristics at the end-point of the limbs almost follow Fechner's low and that there exist the equivalent differences of about 30 % on the magnitude of perceived forces between the steering operation and the pedal operation.

View full abstract

Advanced intelligent vehicles capable of autonomous driving will be commercially available in near future. To build a beneficial relationship between driver and vehicle, it is important to analyze how the drivers would react to autonomous driving compared to conventional human-driven driving. In this study, we develop a driving simulator for analyzing individual driving experience in several road conditions for autonomous and human-driven vehicles. The driving simulator incorporates fundamental functions to connect arbitrary interfaces, create virtual environments consisting of scenarios and events that drivers encounter in real-world driving, and reproduce fully autonomous driving. The results of preliminary experiments showed that experienced drivers opt for human-driven driving mainly due to the flexibility and driving fun it offers, while novices tend to prefer autonomous driving due to its inherent easiness and safety.

View full abstract

In order to provide safe and precise driving assistance for vehicles, a robust road lane detection is needed. This research presents an image processing method based on Hough Transform, line filters to discriminate between road lane and obstacles and vanishing point, extracting road features parameters chronologically and saving them for estimating occluded lanes. This method improved road lane detection comparing to methods using single frames.

View full abstract

The authors are developing "Distributed Power Control System" of "Hard-Link-Type Mobility" using the mechanical coupling device. This mobility is subject to using the Small EV, and make it possible to connect or disconnect the vehicles as necessary. "Hard-Link-Type" has two merits. The first one is robustness in operational environments (the weather, the wireless interference, and etc.). The second one is safety because "Hard-Link-Type Mobility" become a normal tractor when the vehicle control system stop. In addition, the control system enables the mechanical coupling device to avoid the load while the mobility running. This paper describes the improvement points of the control system when assuming practical use.

View full abstract

When a disaster which make many vehicles be left unattended or be stalled would occur, it is difficult to assure driving route for emergency vehicles. In present, however, these vehicles are moved by tow trucks, only one vehicle at a time. Therefore, this study solve the issue by automatic guidance system. This system moves both single vehicles one at a time, and several vehicles simultaneously by a tow truck. This system is easily feasible because it needs no additional devices on board as it focuses change of recent vehicles' structure. Herein, a system which enables remote control of many vehicles left unattended at a time is explained.

View full abstract

In this paper, we propose the automatic parking system using human cooperation. Also, we describe cooperation with parking position indicating device and the in-vehicle camera with pan tilt unit. The in-vehicle camera with pan tilt unit tracks the device and measure that position. Realization of advanced automatic parking system in recent years has a problem to be dependent on the infrastructure of the parking lot. So, not dependent on infrastructure of the parking lot system is required. Than the results of the present experiment, parking position indicating devices and stereo camera of cooperation, has been shown to be effective in the proposed system.

View full abstract

One dangerous condition in driving cars is what is called a split-μ surface, where the friction coefficient of one side of the tyre is different from the other. In such cases even simple braking can result in a spin, and thus several proposals have been made to ensure stability of the vehicle by controlling tractions on each tyres: however, traction control means compromise in stopping distance, and may lead to further accidents. To avoid this problem, in this paper we propose a method to prevent the occurrence of a spin in such conditions through control of vehicle yaw rate with electronic steer-by-wire, with which we aim to minimize the stopping distance as well as the spin behaviour of the vehicle.

View full abstract

In the retail stores and convenience stores, it is important to utilize flow line information which is constructed with a customers' movement locus and their holding time. The flow line information is utilized for estimation of customers' interesting objects and improvement of shelves. However, many conventional customer-data gathering methods are not able to gather the information of customers who do not purchase anything. Some of methods need to attach tags to customers and items. Furthermore, it has to install a lot of sensors densely to get the customers' data which continued from store entry to adjustment. Therefore we propose the novel gathering method of customers' flow line information. To realize our proposed method, we adopt a depth camera and motion capture sensor such as Microsoft Kinect to gather the flow line information. In this paper, we evaluated our proposal customers' flow line information gathering system, and we confirmed that the number of sensors has been reduced.

View full abstract

As a fundamental technology for safety assistance system of PMV, from three-dimensional information of surrounding environment obtained from a stereo camera, the authors propose a method to detect accessible region where walls, steps and obstacles are excluded based on rolling motion of tire. The method obtains three-dimensional point cloud from disparity map calculated with SGMB method at first. Next, rollable region is discriminated based on rolling motion of tire on the point cloud. Then, accessible region is discriminated by excluding rollable region separated from the location of the vehicle by unrollable region. Using the proposed method, an experiment to detect accessible region at a station is carried out. As a result of the verification experiment, the effectiveness of the proposed method is verified.

View full abstract

3D registration is becoming one of the essential technologies for service and industrial robots. For 3D registration, 3D keypoint detection, feature computation on the detected keypoints, and correspondence between model and scene keypoints based on the features are required in most cases. In this research, we will propose high accuracy correspondence method to improve accuracy of 3D registration by using geometric relation among 3D keypoints. As the first step, this paper shows accuracy analysis of the correspondence that depends on 3D shape of a target object and methods to detect 3D keypoints and compute features on the detected keypoints.

View full abstract

For a mobile robot exploring in an unknown static environment, self localization and recognition of target objects are essential problems. In this study, we describe a method to recognize a 3D grid structure using a fixed camera and a RGB-D camera system. Tracking algorithms using a fixed camera are tested along with the Scale Invariant Feature Transform (SIFT) for feature extraction and setting Region of Interest (ROI) with a color marker to improve the efficiency and accuracy of the tracking algorithm. The algorithm of point clouds registration without using feature points matching is also introduced in this work. Finally, the results of recognition of a 3D grid structure with relatively low computational cost is shown experimentally.

View full abstract

This paper proposes a small, lightweight, and easily-relocated sensor terminal named "Portable" for acquiring a variety of environmental information. The Portable is equipped with a variety of sensors including a pyroelectric sensor, a proximity sensor, a sound pressure sensor, a thermometer, a hygrometer, a gas sensor, a flame sensor, and a laser range finder. We introduce three typical applications: abnormality detection, pedestrian tracking and automatic reconfiguration of several Portables for pedestrian tracking.

View full abstract

This paper presents a method to measure leg shape for a fitting simulator to be used in Internet mail order. In the method, a user first takes a picture of the leg from multiple viewpoints. Then feature points are extracted from the images and the corresponding points among images are found. Finally, three-dimensional position of each feature point is calculated based on the correspondence and the camera parameters. However, it is normally difficult to find feature points on legs. So we utilize pattered tights and the feature points are thereby easily obtained, for example, by extracting corners of the pattern. The advantage of the proposed method is that it does not require special equipment and space.

View full abstract

In this paper, the developing of an ultra-small size range image sensor is presented. The sensor is designed for robot hand by using an endoscopy camera and multi-slit laser projector, which can display 3D distance information. The improvements over the previous sensor are size, accessibility and measurement range in close distance. The dimension of our sensor is 16×40×53 mm and only weights about 28 g. By using a universal Webcam chip, the computer can capture the image data without installing the driver. Moreover, the sensor can measure the distance range from 50-150 mm. The effectiveness of the sensor is verified by experiment.

View full abstract

Collision avoidance by braking is effective and is widely used nowadays. Collision avoidance by steering is more effective at high velocity. But the uncertainty of the road surface condition affects the control performance. Therefore, it is very important to know the condition. In this paper, an automatic collision avoidance system is proposed by combination of braking and steering. First of all, a vehicle is braked to estimate the road condition and reduce the velocity. Then, the vehicle is steered to avoid the obstacle by a nonlinear controller with estimated coefficients of road friction.

View full abstract

Electric vehicles are attracting considerable interests, but their shorter mileage per charge is a major drawback. In this study, we focus on the energy loss by slippage between tire and road surfaces, and the minimization of this loss by optimization of the traction/braking force distributions. The electric power consumptions under the running patterns of 10-15 mode and JC08 are theoretically estimated by taking account of the rolling resistance and air drag, utilizing the brush model. The drive-train configurations, such as rear-wheel drive or four-wheel drive will yield different mileage with the same vehicle and tire parameters. In our estimation, the four-wheel drive vehicle equipped with the optimum traction/braking force distribution improves the mileage on low μ road by 4.1% compared to rear-wheel drive vehicle.

View full abstract

We propose the method to detect a stationary obstacle in the environment by a laser range finder and a monocular camera.The method is divided into two systems, the estimation of object system and the image recognition system.The system for estimation of object position uses k- means clustering, and improves disadvantage of k-means clustering that does not guarantee to split clusters optimally depending on the situation.The system for image recognition uses the method of Bag of Features to represent image features and the method of Support Vector Machine in classification.Both systems are run multi-threaded, maintain the real-time properties.Then, we evaluate the performance of each system and confirm behavior when both systems are combined.

View full abstract

Pipe inspection is a critical task. Above all, inspection in the inside of a pipe is usually done using a camera. But, a monocular camera cannot detect the depth and roughness of the pipe. By using a combining camera and a LSA(laser spot array), this kind of information can be detected. However, when using a camera with LSA, it is difficult to cover the whole pipe. To enlarge detecting area, we combine some different pose pictures by detecting orientation and position of the camera by taking a photograph. Finally, we measure in-pipe shapes in three dimensions minutely, making it easily to detect broken points in the pipe.

View full abstract

Conventional three-dimensional measurement methods, such as the passive and active stereo method, are based on triangulation, but the transaction speed tends to be slow and insufficient as a sensor for dynamic mechanical systems and robots. Recently, a technique named Structure Light Field Method was proposed, which reduces a calculation amount for distance estimation by projecting distance information directly to the subject to achieve high-speed measurement of three-dimensional information. This paper reports a method to produce the Structured Light Field using two projectors, and experimental results including the tracking of a moving ball based on the 1000-Hz feedback of the three dimensional information.

View full abstract

Phase shift method is one of the active 3D measurements. In this paper, we consider phase shift with a projector-camera system. Phase shift can reconstruct 3D points with sub pixel accuracy, that is one of main advantages for phase shift compared to the other active 3D measurements. On the other hand, this method assumes linearity between input brightness to a projector and intensity obtained by a camera. Thus measurement accuracy of phase shift depends on optical property for the projector, camera and especially a target object whose optical property is far from linearity. To solve this problem, we will propose high accuracy phase shift method for the target object with nonlinear optical property.

View full abstract

This paper proposes an automatic planning technique of a new laser measurement system using CPS-SLAM named CPS-VIII. The CPS-VIII consists of a parent robot which is equipped with a precise laser scanner and several child robots including wheeled robots and quadcopters. Each robot position is identified by Cooperative Positioning System, CPS. We conducted the 3D modeling experiment and verified that the accuracy of the 3D model is 0.023m after the parent robot moved 270.1m.

View full abstract

The purpose of this study is to develop a target person detecting system using a 3D laser scanner and to propose a efficient grouping method of point cloud data. The system consists of grouping part and target parson determining part. The grouping part performs grouping in the section called plane, and the results are grouped again. The target parson determining part determines whether the target person for grouping result using shape features. The experimental result shows that proposed system can detect a target person as well as existing method and be more quickly than it.

View full abstract

Estimation of 3D location of the human, especially head position and direction estimation is very important and widely used technique for virtual reality. However, a cost of sensor to measure these information is normally quite high. In this paper, an algorithm for estimation of the head direction and position using Kinect is addressed. As a result of the evaluation for estimation, error of head direction estimation is within a 4 degree when the head direction is less than 10 degree.

View full abstract

In recent years, not only industrial robot used in the manufacturing site, the use of a service robot for performing alternate or supplementary work of humans is increasingly robot. So we aim to build the system of mobile service robot quipped with a system that integrates information of external sensor and internal sensor, indoor positioning and map information. We used the IMES to indoor positioning system. In this reports, the experiment result of initial estimation using IMES position and ICP algorithm and autonomous running while providing information using Location information by IMES. And we could show the effectiveness of the system that was created by these experiments.

View full abstract

We had developed the 3D shape measurement method using the robot technology to measure underground pipes.And we got the rough shapes of pipe. Our new aim is the accuracy improvement of the 3D shape measurement.Gyroscope on the robot is used to estimate robot position.Hence the specification of gyroscope is greatly related to accuracy of measurement.In this paper, to improve the accuracy of measurement, we propose a method to choose a gyroscope which can satisfy the target accuracy. We evaluated the measurement error caused by specification of gyroscope by simulation, and we found the index of specification of gyroscope.

View full abstract

In this study, to support handicapped people, especially focusing on a hemiplegic person, a life support system is proposed. For improving their QoL, the system is required to support their early rehabilitation rather than care for whole ADLs. Therefore, we propose a system, which measures and estimates a state or a condition of a patient using multiple RGB-D sensors. In the system, by calculating angle and estimating position of each joints, the Brunnstrom stage of patient is estimated based on a method proposed by Ueda. On the other hand, in the case of measurement using only a RGB-D sensor, some joints are occluded by the other objects and their positions are not able to be estimated. In this paper, to avoid occlusions, a way of measurement a motion of patients using multiple RGB-D sensors is described.

View full abstract

This paper reports the development of small-size manzai robots that are driven by Raspberry Pi controllers. The manzai robots generate their manzai scripts using web news articles based on related keywords given by the audience and the searching results on WWW automatically, and then the robots perform the manzai scripts. Each robot is controlled by distributed RT components executed on the Raspberry Pi controller. The RT components control the manzai robots synchronously. Experimental results show the feasibility of the manzai performance robots.

View full abstract

In this paper, we propose a system framework of a mapping and navigation robot which does not depend on specific robot middleware platforms. Recently, robot middleware have attracted attention widely because that is expected to reduce the cost of robot development. However, many robot middleware platforms have been developed, so it must to define a common framework to design robot system without depending on specific platform. Therefore, we propose an autonomous mobile robot framework which independent on specific robot middleware platforms, hi addition, this model are implemented in RT-Middleware and experimentally evaluated.

View full abstract

For expanding the application of industrial robots, it is necessary that we should develop a robotic system efficiently using modularized hardware and software. The RT-Middleware, RTM, is a platform of the modularizing technology. On the other hand, MOTOMAN industrial robot has open I/F of MotomanSync for operating robot controller from an external PC. hi this paper, we have developed RT-Component. RTC. to control MOTOMAN using RTM and MotomanSync. And we have also built up Pick&Place robotic system in a short time using the RTC. It is useful to reduce development costs of integrated robot system.

View full abstract

Development tools for robot software development should be produced according to a well-structured process to ensure they meet the needs of robot developers and users. The beginning of the process is determining the requirements that the tools must meet. In this article, we describe the first stages in the applying the Volere requirements process to identify the requirements for a software tool used to deploy a component-based robot software application into an execution environment. We describe setting the scope of the tool, determining its context and external interactions, and finding the use cases. Based on these use cases, future work will determine the detailed requirements that a deployment tool should meet.

View full abstract

The conventional 3D-LIDAR was designed for using on robots or vehicles. The 3D-LIDAR is needed to be miniaturized in order to being carried by human for many hours. Therefore, the present author has developed a mobile 3D-LIDAR which has wide view and rotates at high speed. In order to get 3D point cloud of the environment around, it is necessary to synchronize the signal of motor with the signal of scanning laser range sensor. In this way, the author adopts hardware flow control. For the purpose of map generation, scan matching based on occupancy grid maps which is 2D-SLAM is adopted. For the scan matching, 2D point cloud is generated from projection of the 3D point cloud in the specific heights from the floor onto the horizontal plane. Finally, the 3D environmental map is generated by the coordinate transforms of measured 3D point with the obtained sensor position.

View full abstract

RGB-D sensor is a very useful sensor for mobile robot. KINECT v1 is one of the RGB-D sensors. And it is measured the depth information by light cording method. KINECT v1 is developed to use in the room. The light cording method is affected by sunlight. Under the sunlight, this sensor loses the infrared pattern and it cannot calculate the depth information. Then we propose the using the polarized filter to find the infrared pattern. It uses to decrease the reflected light on the digital camera. In some experiments results show that the KINECT v1 is output the depth information under the sunshine.

View full abstract

This paper proposes a stereo camera using two fish-eye cameras. The measurement range of the stereo camera is increased by using fish-eye cameras. Fish-eye images are deformed by using an equirectangular projection to simplify the stereo correspondence search in the images. Affine transformation is applied to images so that feature points of far objects coincide in images of two cameras, in order to remove the mismatches due to the errores of intrinsic and extrinsic parameters. The corresponding points are calculated by using the stereo matching in the deformed image, and the distance is measured from the disparity. The distance measurement accuracy of the constructed stereo camera is evaluated by experiments.

View full abstract

This paper describes autonomous mobile robots navigation system and automatic large-scale 3D-map construction method. In previous works, most of them are using detailed map, which is created by operator in advance. Prior detailed map is a very effective tool to improve working efficiency, stability of the robot, but increasing the workload of operator is undesirable to generalize the robot in the real world. Furthermore, autonomous mobile robots have to function appropriately even in areas where they have never been before. Thus pre-driving details are not needed in our system. LIDAR is mainly used in our system, in creating local map, avoiding the moving object (Motion Planning), and recognizing own position (Localization) and environment around the robot (Environment Recognition). Moreover, we construct the detailed map automatically by using the autonomous robot driving data to improve accuracy of the navigation system without increasing the workload of operator. The availability of proposed method is confirmed by the experiments in Meiji University.

View full abstract

We are currently developing an open-framework for RT-middleware which advances the reusability of RT-component and RT-system. Using the RT-middleware, though the softwares of the robotics elements like actuators, controllers, and sensors, are encapsulated into RT-component which can be easily reused, the way how to collect the source-codes of RTCs is not well-discussed yet. In our open-framework, both a repository management and a build support system are implemented. In this paper, the RTC source-code build-status management system which uses both the repository management and the build support system is shown.

View full abstract

The paper proposes a technique to create a low-cost approximate function for calculation of a large-cost functions in embedded systems. The fast-calculations of arctangent(atan2), sine and cosine functions were developed as its application. The developed atan2 function is 20x faster and 1/3 RAM usage than the standard math library.

View full abstract