Modeling and optimization of implementation aspects in industrial robot coordination

Highlights

• Modeling and optimization of assembly station cycle time in the automotive industry is a key factor.

• The cycle time is also influenced by the robot instructions communicating with a programmable logic controller, since they may introduce delays.

• New mathematical formulation of the problem makes the optimization of the robot programs possible.

• Large instances are handled by a new greedy algorithm showing good results.

• Stud welding and inspection applications show the benefits of the method.

Abstract

Cycle time for an assembly line in the automotive industry is a crucial requirement for the overall production rate and business achievements. In a multi-robot station, besides workload distribution, operations sequencing, and robot motions, an important issue is robot coordination. A key aspect, so far neglected, is the robot motion delay due to the implementation of robot coordination scheduling: our first contribution is to model these delays into a mathematical formulation in order to minimize their contribution to the overall cycle time. Two different setup scenarios, based on typical hardware setups, are modeled and optimized. The major contribution is an efficient heuristic algorithm to optimize our suggested model, which provides near optimal solutions in fraction of the time required by a general optimization package. We have also integrated such algorithm in a CAM software for robot offline simulation to be able to run industrial scenarios with real CAD models from the automotive industry. The results show that it is possible to reduce the delays due to the synchronization points and consequently cycle time, while still maintaining the pre-computed nominal time scheduling for robot coordination.

Introduction

The assembly line in the automotive industry usually consists of several automated cells where robots perform sealing and welding operations on the Body in White (BiW). Sensors, robots, and other equipment are usually coupled to a Programmable Logic Controller (PLC) that decides the timing of operations based on its programmed logic behavior. Besides moving the BiW, clamping, and controlling safety operations, PLCs are often used to schedule robot programs such that the robots do not collide with each other, see Fig. 1. This is done by signaling each robot whether it can continue its program or not, based on their occupancy of shared workspace volumes. The main rule is that only one robot at a time can occupy a common workspace, preventing any potential collision with other robots. This is a very critical task since the line throughput is heavily influenced by that and since incorrect behavior may result in robot damages, line stops, with all the corresponding drawbacks.

Therefore, an optimized schedule of robot operations is crucial and off-line simulation of robot programs plays a central role to achieve that. Indeed, simulation tools modeling the reality in a virtual environment are becoming a key factor for success in the very tough automotive market. Besides modeling and simulation, the most important feature in a simulation tool is the possibility to automatically or semi-automatically optimize the process.

We focus on this optimization aspect, applied to the minimization of cycle time in industrial robot stations, especially automotive, see [1], [2], [3]. In previous works, robot coordination has been achieved by changing the start time of each operation or robot motion, see [4], [5]: the nominal duration of each motion has been considered independent of its time scheduling. After this has been done in the virtual environment, the simulation engineer needs to transfer the obtained solution into the real world.

However, in reality, different hardware setups are adopted: robots communicate directly with each other through the robot controllers digital input/output, or they communicate through a PLC. Moreover, robot manufacturers might implement in different ways the instructions for signal exchanging. At the same strength, automotive industries can program differently how the handshaking between PLC and robot controllers behave. It turns out that these details have a large influence on the real cycle time: an aspect that has not been modeled and considered during the optimization, so far. Indeed, the robot might stop its motion to perform the handshake, increasing in a non negligible way the overall pre-computed cycle time. This communication aspect is becoming much more evident in Human-Robot applications where common tasks can be performed and must be coordinated in a safe way, see [6].

To resolve this neglected problem, we model the time delays related to the hardware specifications and consider them in a post processing step to minimize their influence on the cycle time.

This article is organized in the following way: first, a description and a mathematical model is given for the path coordination problem, in Section 2. Then the main problem is described in Section 3 where we highlight the importance to minimize the number of synchronization via points. Our contributions are presented in Sections 3.1 and 3.2, with different mathematical programming formulations, and in Section 4, with a new algorithm aiming to find near optimal solutions. In the last section computational experience and results are presented, based on two industrial scenarios. Finally, a conclusion section summarizes the results with some suggestions about future work.