Laser Scanning Based Object Detection to Realize Digital Blank Shadows for Autonomous Process Planning in Machining
Round 1
Reviewer 1 Report
This paper presents an approach for measuring a blank prior to machining using a laser line scanner. It addresses an interesting topic, which represents a part on the way to autonomous manufacturing. Unfortunately, the description lacks a critical discussion of the results. The authors refer several times to the "digital twin (CAD file of the raw part)". But is the CAD file or also the measured raw part geometry of the component already a "digital twin"? The publication, e.g., presented by Bergs et al. ("The Concept of Digital Twin and Digital Shadow in Manufacturing", Procedia CIRP 101 (2021) 81-84) should be considered.
On p. 4 it is described: "Based on the digital blank twin, the whole milling process for the defined workpiece is planned automatically by an embedded CAD/CAM planning". Unfortunately, the approach behind this is not addressed in the publication. Completely automated CAM planning - especially without any restrictions regarding the complexity of the component - appears to be very complex. Here it should be clearly worked out whether this has already been implemented or represents future work. The complexity should also be critically discussed in the paper. The addressed automatic definition of the milling strategy incl. process parameter selection should also be described in more detail. In particular, this should be illustrated with examples.
In their summary, the authors state that "an approach for a fully automatic workpiece detection was proposed" (p. 14). This statement is very general, since it was shown on two simple examples (cuboid and cylinder). Furthermore, the authors themselves restrict these statements later on by stating that this approach can only be used for "a rough blank detection" (p. 14) due to its accuracy. This should be critically discussed already at the beginning. The authors also state "If there are higher requirements, a combination of the introduced approach and a touch probe can be considered" (p. 14). Here, it should be discussed in detail whether the assumption of an ideal geometry (cuboid or cylinder) in combination with a touch probe would not be also useful and also faster. In the discussion it is listed that "different shapes [...] could automatically be determined." (p. 14). Since only two simple ruled geometries were used here, this statement should be reconsidered.
Based on the above comments, I do not recommend the submission in its present form for publication in this journal.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The manuscript addresses an interesting topics and it is well-written and presented. In the reviewer's opinion the manuscript can be accepted after addressing the issues provided in the following comments:
1) Manuscript should be proofread to fix language mistakes. For example in pag.2. "The exact ... as possible [4]. After ... as possible." authors report two different sentences with the same meaning. In pag. 13 "The results show, that .." comma should be suppressed, and "These deviations could have different influences: First ..." the sentence should be revised.
2) In section 4, a photo of the experimental set-up (machine tool, scanning system) should be included.
3) In Figure 2 the optional possibility of combining the proposed approach with the automatic use of a touch probe to refine the results could be included.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
In the revised version, the authors clarified most of the reviewer’s remarks. Nevertheless, there are still a few open questions:
- In chapter 3, the authors mention “This work aims to investigate a fully automatic method for blank detection, which is integrated in an overall developed CAD/CAM chain for an autonomous milling process.”. The paper focusses on the first part (blank detection). The “autonomous milling process” is not shown. It is therefore recommended that these statements should be revised in the paper, e.g., by “which could be integrated in an ….”. The same applies to the description on p. 4.
- In chapter 2, it is mentioned: “Furthermore, probing is a time-consuming process that can take up to several hours.” (p. 2). Examples of components for which the measurement takes so long should be given here. A probing of simple geometries, which are addressed in this paper, is assumed to be faster.
- On p. 6, it is described “After generating an NC code for the sensor path with CAD/CAM software”. A detailed description of how this is done automatically is desired.
- The paper needs an overall linguistic revision.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf