Add pages related to ARPA-E Mapping and confined-space SLAM (#35)

* Add page confined-space_slam * Add many pipe inspection related pages * Finish adding contents * Revert last change, delete unused files, add more tags
biorobotics · Aug 25, 2023 · d6b8b11 · d6b8b11
1 parent 42a7396
commit d6b8b11
Show file tree

Hide file tree

Showing 29 changed files with 136 additions and 2 deletions.
diff --git a/_posts/2023-08-15-autonomous-needle-insertion.md b/_posts/2023-08-15-autonomous-needle-insertion.md
@@ -5,6 +5,8 @@ categories:
 description: "Automatic Needle Insertion for Vessel Access"
 published: true
 image: img/posts/AutoNeedleInsertionforVesselAccess.png
+tags: 
+    - medical_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-15-ultrasound-ai.md b/_posts/2023-08-15-ultrasound-ai.md
@@ -5,6 +5,8 @@ categories:
 description: "Toward Practical Ultrasound AI Across Real-World Patient Diversity"
 published: true
 image: img/posts/image100.png
+tags: 
+    - medical_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-16-power-plant-application.md b/_posts/2023-08-16-power-plant-application.md
@@ -5,6 +5,8 @@ description: "Autonomous Force-Controlled Ultrasound Scanning"
 categories:
   - applications
 image: img/posts/powerplant.png
+tags:
+  - inspection_and_repair
 ---
 Current power plant and industrial inspection tools are limited in their ability to access difficult-to-reach locations. Some systems like inspection crawlers can easily drive towards a destination, but cannot fit through small openings or navigate tight turns. Other systems like borescopes are flexible and can fit through small cross-sectional access points, but are limited in their ability to push through turns or up vertical sections of piping. Snake robots combine the best of both types of tools: a small cross-section, flexibility to navigate bends, and the ability to locomote and even climb vertically. This enables the snake robots to reach locations that are otherwise inaccessible. The robots are ideal for performing remote visual inspection (RVI) or carrying a sensor payload to conduct nondestructive 
 evaluations (NDE).

diff --git a/_posts/2023-08-17-body-slam.md b/_posts/2023-08-17-body-slam.md
@@ -6,6 +6,7 @@ title: Body SLAM
 categories: research
 tags:
   - medical_robotics
+  - simultaneous_localization_and_mapping
 ---
 
 <!-- <div class="column is-12"> -->

diff --git a/_posts/2023-08-17-nuclear-inspection.md b/_posts/2023-08-17-nuclear-inspection.md
@@ -5,6 +5,7 @@ published: true
 title: Nuclear Inspection
 categories: applications
 tags:
+  - inspection_and_repair
   - expeditionary_robotics
 ---
 Certain areas of nuclear power plants or nuclear waste storage locations are highly radioactive and cannot be safely accessed by people. While a variety of conventional robots have been developed to help provide access to radioactive locations, there are still limitations to where these robots can access and move. Snakelike robots' size and maneuverability could provide access to otherwise unreachable locations within these radioactive environments. In particular, these robots could address the challenge of inspecting inside of dry storage casks, waste storage tanks, piping within nuclear power plants or waste vitrification plants, or in other radioactive locations. The robot's ability to swim could also enable them to perform underwater tasks such as spent fuel pool inspection or reaching locations within boiling water reactors.

diff --git a/_posts/2023-08-21-e-waste-recycling.md b/_posts/2023-08-21-e-waste-recycling.md
@@ -4,6 +4,8 @@ date: '2023-08-21 15:16 -0400'
 published: true
 title: E-waste Recycling
 categories: research
+tags:
+  - recycling_robotics
 ---
 We live in a world where most sensitive information is stored on our devices. Eventually, like many things, those devices are thrown away. Existing waste management has more than half a century of experience in handling materials the right way, but dealing with electronics items poses a new challenge and if not done properly can be hazardous to the environment. The main idea is to identify parts that can be remarketed or recycled. In this work, we are trying to classify electronics, develop semantic understanding of their components using X-ray images, manipulate them to respective stations and then autonomously disassemble them to smaller components. 
 

diff --git a/_posts/2023-08-21-modular-robots.md b/_posts/2023-08-21-modular-robots.md
@@ -7,6 +7,7 @@ published: true
 image: img/posts/eigenbot.gif
 tags:
     - locomoting_systems
+    - modular_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-21-modular_snake.md b/_posts/2023-08-21-modular_snake.md
@@ -4,7 +4,8 @@ date: '2023-08-17 20:33 -0400'
 published: true
 title: Modular Snake
 categories: robots
-tags:
+tags: 
+    - modular_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-21-sea_snake.md b/_posts/2023-08-21-sea_snake.md
@@ -5,6 +5,8 @@ published: true
 title: SEA Snake
 categories: robots
 tags:
+    - modular_robotics
+    - locomoting_systems
 ---
 
 The **series-elastic actuated snake robot (SEA-snake)** is a modular snake robot composed of any number of modules arranged in a chain. The head typically carries a sensor package including a camera with LED lights. This robot is used for research projects on compliant control and mobility, especially in unstructured terrain or confined spaces. Its two-inch diameter profile, and ability to wrap around poles, allows it to move through both the outsides and insides of pipe networks.

diff --git a/_posts/2023-08-21-snake-robot_systems.md b/_posts/2023-08-21-snake-robot_systems.md
@@ -7,6 +7,7 @@ published: true
 image: img/posts/reu_system.png
 tags:
     - locomoting_systems
+    - modular_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-21-snake-robots.md b/_posts/2023-08-21-snake-robots.md
@@ -7,6 +7,7 @@ published: true
 image: img/posts/sidewinding.gif
 tags:
     - locomoting_systems
+    - modular_robotics
 ---
 
 <figure>

diff --git a/_posts/2023-08-23-confined-space_slam.md b/_posts/2023-08-23-confined-space_slam.md
@@ -0,0 +1,29 @@
+---
+image: img/posts/blaser-square.png
+categories:
+    - research
+date: '2023-08-23 16:29 -0400'
+published: true
+title: 'Confined-Space Sensing and 3D Reconstruction'
+tags:
+    - confined_space_robotics
+    - simultaneous_localization_and_mapping
+---
+Confined-space applications require specialized 3D sensors to scan the environment, detect objects, and avoid obstacles. Compared to mainstream sensors being widely used among the robotics community, such as Lidar and machine vision cameras, 3D sensors for confined spaces have additional requirements on the miniature size, ultra-short sensing ranges, and high accuracy.
+
+There are no COTS solutions available to fulfill the ultra-short-range but high-precision 3D perception challenge. Most of the commercially available sensor systems are bulky, “farsighted”, and require external computers to perform 3D processing. It would be difficult to even obtain and let alone miniaturize an integrated sensor system, and through our work, we have developed a deep insight into creating these sensors and intelligently processing information from these sensors.
+
+### Ultra-Compact Sensor for High-Accuracy Close-Up 3D Reconstruction
+
+<figure>
+ <img src="img/posts/blaser-handheld.png" width="55%"/>
+</figure>
+
+We have developed several working prototypes of miniature 3D sensors which are well-suited to confined-space manufacturing and inspection applications. The foundamental and enabling technology behind these sensors is a multi-sensor fusion simultaneous localization and mapping (SLAM) framework. It integrates a camera, a laser profiler, and an inertial measurement unit (IMU) into an ultra-compact sensor package capable of performing real-time, dense, and colorized 3D reconstruction of close-range objects with sub-milimeter-grade scanning accuracy. We are actively working on improving the hardware performance, software efficiency, and integrability with other robotic systems, with the goal to make it the best option for short-range industrial robotic perception solutions for all applications.
+
+<figure>
+ <img src="img/posts/blaser-scanning.png" width="70%"/>
+</figure>
+
+**Example use case: Industrial Part 3D Scanning**
+<figure class="image is-16by9"><iframe class="has-ratio" src="https://www.youtube.com/embed/LECQlbpoq1g" frameborder="0" allowfullscreen></iframe></figure>
diff --git a/_posts/2023-08-23-pipe-inspection-application.md b/_posts/2023-08-23-pipe-inspection-application.md
@@ -0,0 +1,37 @@
+---
+published: true
+title: "Pipe Inspection"
+categories:
+  - applications
+image: img/posts/pipe-inspection-application-cover.png
+tags:
+  - inspection_and_repair
+  - confined_space_robotics
+---
+
+
+Pipelines, crucial infrastructures supporting human civilization, may experience degradation from various factors like corrosion, geological subsidence, and improper plumbing or digging, leading to economic losses and hazardous incidents. The inspection and maintenance of pipelines are of paramount importance. Conventionally, nondestructive testing (NDT) and inspection for pipes often involves the use of push-rod borescopes, which typically consists only of a camera for 2D video footage capturing. Due to the monomodal nature of 2D images, such methods may fall short of the objectives to detect and localize anomalies. 
+
+Leveraging the [In-pipe Inspection Sensoring]({% post_url 2023-08-23-pipe_inspection_sensors %}) and [SLAM]({% post_url 2023-08-23-pipe_slam %}) technology developed at our lab, we are capable of acquiring a more comprehensive digital record of the pipe interior through automatically collecting sensory data and perform RGB-D reconstructions in pipes, providing a combination of visual, 3D, and georeferencing information. Conbined with AI-based anomaly detection and cloud computing technology, we aim to use the in-pipe data for more sophisticated pipe condition monitoring, assessment, and anomaly localization.
+
+Our objective is to provide intelligent and efficient trenchless rehabilitation solutions that deliver the highest level of precision and insight with minimal human involvement. By revolutionizing the way pipelines are inspected and maintained, we aim to promote safety, reliability, and environmental sustainability across the industry.
+
+
+**To learn more about our in-pipe inspection technology, visit [this website](pipe.report).**
+
+### Wide Ranges of Use Cases Across Industries
+
+Relevant industries include but are not limited to natural gas, infrustructure (conduits), sewer, and wind turbine.
+
+### Cloud-Based Data Storage and Web-Based Visualization
+
+To facilitate downstream usage, we are developing a in-pipe scanning database and user interfaces for interactive visualizating the scans. [This website](vis.pipe.report) shows a reconstructed point cloud sample of the pipe interior.
+
+### Data Capturing for Anomaly Detection and Predictive Maintenance
+
+After capturing the visual images and 3D point clouds, it is essential to analyze the data and pinpoint the existing anomalies. Furthermore, we also envision the capability to predict future anomalies, so that a maintenance plan can be preemptively outlined and the defects can be prevented in advance. 
+
+<figure>
+ <img src="img/posts/pipe-anomalies.png" width="100%"/>
+</figure>
+
diff --git a/_posts/2023-08-23-pipe_inspection_sensors.md b/_posts/2023-08-23-pipe_inspection_sensors.md
@@ -0,0 +1,19 @@
+---
+image: img/posts/pipe-blaser-v5.gif
+categories: robots
+date: '2023-08-23 16:29 -0400'
+published: true
+title: In-Pipe Inspection Sensor Suites
+tags:
+  - confined_space_robotics
+  - inspection_and_repair
+---
+
+Various prototypes of in-pipe inspection sensor suites have been developed to perform data acquisition and SLAM in pipes with different diameters, from 12'' or more down to 4''. The sensor suite family surpasses conventional borescopes with its advanced capabilities in [low-drift dense RGBD reconstruction]({% post_url 2023-08-23-pipe_slam %}), AI-based anomaly detection, and embedded computing technology. The sensor packages are attachable to existing actuated in-pipe crawler robots and mobility platforms.
+
+
+<figure>
+ <img src="img/posts/pipe-robots-3.png" width="100%"/>
+</figure>
+
+Planned efforts involve further scaling down the sensor package size so that they can be deployed in narrower pipes with a diameter of 2'' or less, while maintaining the localization and mapping capability. The challenge to build smaller sensor suites primarily lies in the tradeoff between size and onboard computational resource.
diff --git a/_posts/2023-08-23-pipe_slam.md b/_posts/2023-08-23-pipe_slam.md
@@ -0,0 +1,33 @@
+---
+published: true
+title: "Pipe SLAM"
+categories:
+  - research
+image: img/posts/pipe-mapping-gif.gif
+tags:
+    - confined_space_robotics
+    - simultaneous_localization_and_mapping
+---
+
+
+For [Pipe Inspection]({% post_url 2023-08-23-pipe-inspection-application %}), simultaneous localization and mapping (SLAM) can be used to reconstruct the pipe’s inner surface, providing a more comprehensive digital record of the pipes compared to conventional vision-only inspection. However, in narrow pipes, the sensing hardware and software used in most conventional methods suffer low localization and mapping accuracy. In terms of hardware, some methods are entirely unable to operate inside small pipes due to the limitation of minimal sensing range and bulky sensors. Algorithmically, the slower sensor motion in confined spaces can lead to insufficient IMU excitation and thus an incorrect metric scale or IMU bias initial estimation. The lack of visual and geometric features also poses a significant challenge to state estimation. Moreover, in environments where global positioning information is unavailable, any SLAM algorithm that only relies on spatially or temporally local measurements is likely to experience amplified odometry drift over the long run due to the accumulation of uncorrected dead-reckoning errors.
+
+<figure>
+ <img src="img/posts/pipe-slam-flow.png" width="70%"/>
+</figure>
+
+Built on top of our [Confined-Space Sensing]({% post_url 2023-08-23-confined-space_slam %}) technology, we have developed an in-pipe SLAM solution that fuses a monocular camera (V), an inertial sensor (I), a ring-shaped laser profiler (L), and an additional Lidar (L) into a compact sensor package optimized for in-pipe operations. Our tightly-coupled sliding-window-based SLAM pipeline is fused with a combination of Lidar-based constraints derived from pipe geometric structure for long-term drift reduction. Real-world experiments in 12''-diameter natural gas pipes have shown our algorithm's advantage compared to the performance of other state-of-the-art algorithms both in terms of long-distance localization accuracy and mapping quality.
+
+Our recent efforts involve testing the sensor hardware and software in more diverse environments, which includes multiple pipe junctions and bendings. We are also building various [In-Pipe Inspection Sensor Suite]({% post_url 2023-08-23-pipe_inspection_sensors %}) versions for different pipe diameters ranging from 12'' or more down to 2''. Meanwhile, we are also working on enabling loop-closure and adding hybrid topological-metric maps for larger-scale pipeline mapping and map abstraction.
+
+<figure>
+ <img src="img/posts/pipe-real-time.gif" width="70%"/>
+</figure>
+
+<figure>
+ <img src="img/posts/pipe-mapping-side.gif" width="70%"/>
+</figure>
+
+<figure>
+ <img src="img/posts/pipe-close-up.png" width="70%"/>
+</figure>
diff --git a/img/posts/blaser-handheld.png b/img/posts/blaser-handheld.png
diff --git a/img/posts/blaser-scanning.png b/img/posts/blaser-scanning.png
diff --git a/img/posts/blaser-square.png b/img/posts/blaser-square.png
diff --git a/img/posts/pipe-anomalies.png b/img/posts/pipe-anomalies.png
diff --git a/img/posts/pipe-blaser-v3.png b/img/posts/pipe-blaser-v3.png
diff --git a/img/posts/pipe-blaser-v5.gif b/img/posts/pipe-blaser-v5.gif
diff --git a/img/posts/pipe-close-up.png b/img/posts/pipe-close-up.png
diff --git a/img/posts/pipe-inspection-application-cover.png b/img/posts/pipe-inspection-application-cover.png
diff --git a/img/posts/pipe-mapping-gif.gif b/img/posts/pipe-mapping-gif.gif
diff --git a/img/posts/pipe-mapping-side.gif b/img/posts/pipe-mapping-side.gif
diff --git a/img/posts/pipe-real-time.gif b/img/posts/pipe-real-time.gif
diff --git a/img/posts/pipe-robots-3.png b/img/posts/pipe-robots-3.png
diff --git a/img/posts/pipe-slam-flow.png b/img/posts/pipe-slam-flow.png
diff --git a/index.md b/index.md
@@ -18,7 +18,7 @@ pis:
       name: Matthew Travers
 ---
 
-# Our lab
+# Our Lab
 
 The Biorobotics Lab reduces complicated high-dimensional problems found in robotics to low-dimensional simpler ones for design, analysis, and planning. Often, we look to biology for inspiration and sometimes, we return the favor by providing analysis that models biology.