1. Abstract

The concept of the vehicle has changed with technological innovations on last decade. Today we can call these changes basically as “CASE” (Connected, Autonomous/Automated, Shared, and Electric). The ease of product access on the user side and the mass production related works have increased worldwide production volumes. This issue has resulted in a greater demand for manpower in the sector. In addition, management, productivity and profitability related difficulties have occurred. In this project, improvements were made mainly in the area of the productivity through the automation of “vehicle transfer operations in plant operations”, which is one of a major problem and a manpower/hour consuming task. This system named as Remote Control Auto Driving System (RCD). The advance technology used system enabling unmanned, secured operations, were implemented in mass production environment earlier than the rest of the world (Fig.1).

Fig.1　Unmanned vehicle transportation by RCD.

2. Technology

2.1 System Overview

A control system performs functions of perception, judgement and vehicle move operations. Control system and the vehicle communicate via wireless communication. Finally, outside located control system manage almost all vehicle actions until the end point. Figure 2 shows the control flow. (1) Vehicle location estimation is performed by real-time with image processing after sensing by mainly infrastructure cameras. (2) The vehicle motion control instructions are comprehensively calculated by the control system based on position information, target route, vehicle status, etc. (3) Transmits motion control commands to the vehicle via wireless communication. (4) The vehicle runs autonomously by activating many systems in same time such as the powertrain, steering and brakes according to the received motion control instructions. The control system run mentioned steps with low latency and high-speed cycles to enable accurate vehicle motion control.

Fig.2　RCD control flow.

2.2 Vehicle positioning accuracy & cost reduction

The position coordinates and azimuth angle are calculated by vehicle positioning dynamically. In addition, motion control related high accuracy location information is required. Existing tech., use in-vehicle external sensors or LiDAR for positioning. However, cost is an issue in both cases and has been an obstacle to their widespread use. RCD uses advanced technology infrastructure together with cameras and an image processing model that can perform object detection and segmentation simultaneously with high speed. This enables fast and accurate estimation in milliseconds from the vehicle’s outline. Finally, accurate calculation and localization could be implemented successfully.

2.3 Calculation of accurate motion control

The purpose of motion control is to follow the target route by the vehicle autonomously. The main challenges about the control: Communication and processing delays. In addition, the effect of a deteriorating SN ratio of the positioning information, which cannot be completely avoided in image processing. We have implemented two countermeasures to eliminate these obstacles. The first is to adopt a two-degree-of-freedom control configuration that combines feedforward control (FF) and feedback control (FB). The FF determines the angle of steering gear operation from the curvature of the target path. The FB derives the steering angle from the lateral deviation of the vehicle position and the target path. To ensure stability, The FF is used to follow the path while suppressing the dependency on FB. (Fig.3)．

Fig.3　Target steering angle. calculation

The second countermeasure is an accuracy improvement method that does not rely solely on image processing to estimate positioning information, but also uses vehicle sensors. The image processing has the advantage of acquiring absolute values, but it also has the disadvantage of worsening the SN ratio. On the other hand, vehicle-mounted sensors have a good SN ratio, but they have a problem of accumulating related integration errors. Therefore, the accuracy of the position estimation was improved by utilizing the merits of both image processing and onboard sensor integration. As a result of this study autonomous movement achieved successfully.

2.4 Maintain of wireless communication

Communication quality is very important, because control values must be transmitted and received with high-speed cycles via wireless communication. However, wireless communication has the risk of quality degradation due to communication interference, noise and fading. Therefore, RCD has introduced redundant wireless communication. This is a logic that constantly evaluates indicators related to communication quality, such as radio wave strength. In addition, adopts communication paths that can be judged to be of high quality overall (Fig.4). This greatly reduced the probability of communication delays and interruptions. In the event of a communication black out, the vehicle detects it, and automatically applies the brakes and stops. Finally, in case of any failure system stops the autonomous driving control and ensure safety conditions.

Fig.4　Wireless redundancy.

3. Conclusion

RCD is an equipment-driven vehicle technology with minimal requirements for vehicle and functionality. Easy to use products and new tools are creating about easy deployment. Therefore, it has the potential to be used not only to improve productivity in factories, but also to enhance safety, security and convenience in a variety of services.

On the other hand, since RCD consists of a combination of various technologies related to CASE, continuous evolution is necessary to utilize RCD with more conditions. Developments and improvements will be continued at real manufacturing environment and will be promoted technological innovations that contribute to “freedom of movement for all”.

Takuro Sawano
Member，TOYOTA MOTOR CORPORATION（1, Motomachi, Toyota-shi, Aichi）

Takeshi Kanou
TOYOTA MOTOR CORPORATION（1, Toyota-cho, Toyota-shi, Aichi）

Shogo Yasuyama
TOYOTA MOTOR CORPORATION（1, Motomachi, Toyota-shi, Aichi）

Kento Iwahori
TOYOTA MOTOR CORPORATION（1, Motomachi, Toyota-shi, Aichi）

KeigoIkeda
Member，TOYOTA MOTOR CORPORATION（1, Motomachi, Toyota-shi, Aichi）