logseq/pages/hls__RoSE2024_paper_4_1704401083677_0.md

file:: RoSE2024_paper_4_1704401083677_0.pdf file-path:: ../assets/RoSE2024_paper_4_1704401083677_0.pdf

• The growing complexity of work cells necessitates the development of improved techniques, methodologies, and tools for their creation, optimization, and debugging. ls-type:: annotation hl-page:: 1 hl-color:: green id:: 659d4089-7078-43e8-998b-6dc36ec83968 hl-stamp:: 1704804492761
• application of dynamic visualizations for the debugging process using domainspecific knowledge ls-type:: annotation hl-page:: 1 hl-color:: purple id:: 659d4094-be64-4958-9a18-7e3a4fb19edb hl-stamp:: 1704804502339
• These visualizations are tailored for debugging collaborative robots, focusing on pick-and-place applications, and are integrated into a proof-of-concept tool. ls-type:: annotation hl-page:: 1 hl-color:: purple id:: 659d40be-c5e8-4aa6-be74-a2182a4e9d5c
• we showcase its ability to enable operators to verify the correctness of the robot’s behavior and identify program failures using several case studies. ls-type:: annotation hl-page:: 1 hl-color:: purple id:: 659d40d6-ef97-427b-b234-4b08df4ba442
• While the figure tends to vary, a general agreement exists that 60% to 80% of a system’s budget is spent on software maintenance [4]. ls-type:: annotation hl-page:: 1 hl-color:: green id:: 659d4137-6198-4c9e-b921-a491cca39e06 hl-stamp:: 1704804666840
• despite the established significance of debugging practices in software engineering, their integration into the robotics domain still needs to be explored. ls-type:: annotation hl-page:: 1 hl-color:: purple id:: 659d4e2c-ed4c-4a6d-b94d-bc291d0ece18 hl-stamp:: 1704807982914
• In this study, we introduce a visualization tool designed for debugging collaborative robots, specifically addressing common challenges in pick-and-place applications involving robotic arms. ls-type:: annotation hl-page:: 1 hl-color:: purple id:: 659d4e57-4484-4ce8-9a4d-5f1f160634f9
• The tool presented in this paper differs from previous approaches by presenting runtime information in state resolution, intricately linking the program and its contents to execution. Consequently, this facilitates the presentation of visualizations and pertinent debugging metrics within a more contextually enriched execution framework ls-type:: annotation hl-page:: 1 hl-color:: yellow id:: 659d4ed4-1175-4b5e-84d8-1af0a9cc3aa5 hl-stamp:: 1704808152918
• facilitating the visualization of diverse sensor and state data from robots, commonly used in the pre-deployment phase. ls-type:: annotation hl-page:: 1 hl-color:: green id:: 659d4f50-a944-4126-a1a6-9d5d03e5aa3f hl-stamp:: 1704808275093
• the visualizations crafted in our tool apply to cobots operating outside of the ROS domain. ls-type:: annotation hl-page:: 1 hl-color:: yellow id:: 659d4f62-598c-4e63-a9cc-dfecd1be58e8 hl-stamp:: 1704808294137
• takes this concept a step further by enabling for monitoring strategies, i.e., configuring alarm systems to reduce unforeseen halts. ls-type:: annotation hl-page:: 2 hl-color:: yellow id:: 659d4fbb-a901-482d-b8bc-d0781432cf18
• izRob lacks the seamless integration between event information and runtime ls-type:: annotation hl-page:: 2 hl-color:: green id:: 659d4fea-d2c4-444e-b04c-be713be0a889 hl-stamp:: 1704808429007
• various initiatives explore the application of augmented or virtual reality (AR or VR) for visualizing data to comprehend robot behaviors ls-type:: annotation hl-page:: 2 hl-color:: green id:: 659d4ff7-7537-41e1-9c7e-a5fef572fd45
• RELATED WORK ls-type:: annotation hl-page:: 1 hl-color:: yellow id:: 659d5029-2949-420a-a88a-e088de50b2fa hl-stamp:: 1704808494093
• Even with partial grasping, the behavior is correct, though potentially less desirable for operators due to reductions in the vacuum gripper’s effectiveness. ls-type:: annotation hl-page:: 2 hl-color:: green id:: 659d5106-ca2c-422b-8dad-1fa785fa0417 hl-stamp:: 1704808712378
• Operators armed with comprehensive information can understand and adjust the arm’s behavior. However, current cobot application evaluations rely on continuous visual inspection, and new deployments are typically tested across an entire work shift. ls-type:: annotation hl-page:: 2 hl-color:: green id:: 659d5125-48e7-43ac-bb93-1d1d93e808c5
• Our visualizations utilize data from the cobot and its environment, enabling operators to assess failures remotely ls-type:: annotation hl-page:: 2 hl-color:: yellow id:: 659d514e-eac6-4219-b31c-50bb5d4ac425 hl-stamp:: 1704808832318
• rrors were detected in 40% of the recorded cycles ls-type:: annotation hl-page:: 3 hl-color:: yellow id:: 659d5274-92f5-47e9-9bee-74d420763c28
• triangular figures ls-type:: annotation hl-page:: 3 hl-color:: green id:: 659d52f8-60ef-47d5-a252-8dcf7dad2498 hl-stamp:: 1704809210053
• previous ls-type:: annotation hl-page:: 3 hl-color:: green id:: 659d52fc-6517-4adc-a807-d29e7fd9542b
• circles ls-type:: annotation hl-page:: 3 hl-color:: blue id:: 659d5301-4165-4086-bd82-1395d8f7b5cb hl-stamp:: 1704809219535
• current ls-type:: annotation hl-page:: 3 hl-color:: blue id:: 659d5306-f7ef-49c3-af67-9c9b32fface5
• squares ls-type:: annotation hl-page:: 3 hl-color:: purple id:: 659d5309-0170-4fd6-b159-4000715eebb7 hl-stamp:: 1704809228064
• following ls-type:: annotation hl-page:: 3 hl-color:: purple id:: 659d530d-3a9a-4b98-aa94-82c4c1eea3c6
• future ls-type:: annotation hl-page:: 3 hl-color:: purple id:: 659d530e-a576-4046-abb3-13ff836a6499
• The tool incorporates multiple visualizations that aid users in discerning the success or failure of the robotic arm in picking up objects ls-type:: annotation hl-page:: 4 hl-color:: purple id:: 659d5357-6d24-4f33-9ab3-db096aeeb2b6
• we demonstrate the tool’s efficacy in enabling users to comprehend the behavior of robotic arms ls-type:: annotation hl-page:: 4 hl-color:: purple id:: 659d535e-4d88-4076-b370-7a5315a5fb15
• 412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464raw variable names Mouse-hover shows values at step Both suction cup signals above configured target One suction cup signal above target One suction cup signal below target Both suction cup signals below target Controller steps Percentage of gripper max vacuum level(b) Partial (incorrect) object grasping(c) Failed object grasping(a) Successful object grasping(d) Variable overview and step selector Values of variables created in script Note, these variables were also used for the visualizations Values at selected step Figure 3: Vacuum level visualization for success, fail and partial grasping. Explanatory text in red. emphasizing pick-and-place applications. The tool incorporates multiple visualizations that aid users in discerning the success or failure of the robotic arm in picking up objects. Through an experimental setup, we demonstrate the tool’s efficacy in enabling users to comprehend the behavior of robotic arms. Users can swiftly assess whether the robotic arm aligns with the intended behavior specified in the source code. However, it is essential to note that our findings are indicative, and further validation through user involvement is imperative to substantiate the tool’s effectiveness in cobot applications debugging. We aim to explore avenues for extending this tool to debug a broader spectrum of cobot applications effectively. Therefore, in future work, our focus will entail a comprehensive analysis of diverse cobot applications, en ls-type:: annotation hl-page:: 4 hl-color:: purple id:: 659d5369-b6e3-41fe-afa6-f32db23f4543