Tag: computer vision

Categories
Applied Innovation

Leveraging AI, ML, CV, and NLP to transform unstructured data into valuable intelligence

3D Digital Twins: The Key to a More Efficient, Safer, and Sustainable Future
Categories
Applied Innovation

Leveraging AI, ML, CV, and NLP to transform unstructured data into valuable intelligence

In today’s digital era, organizations are swimming in a vast ocean of data, with a significant portion of it residing in unstructured documents. These documents, such as emails, contracts, research papers, and customer feedback, hold a wealth of valuable information waiting to be unlocked. However, extracting meaningful insights from this unstructured data has traditionally been a daunting task. Enter the power of Artificial Intelligence (AI), Machine Learning (ML), and Natural Language Processing (NLP). These transformative technologies are revolutionizing the way businesses derive value from the data encapsulated within unstructured documents.

Unstructured documents differ from structured data sources, such as databases or spreadsheets, as they lack a predefined format or organized data model. They contain free-form text, images, tables, and diverse information types, making them challenging to analyze using conventional methods. However, advancements in AI, ML, and NLP have paved the way for extracting valuable insights, patterns, and knowledge from these untapped resources.

By applying intelligent algorithms and techniques, businesses can gain a competitive edge, drive innovation, and make informed decisions based on comprehensive data analysis. NLP techniques enable the classification of unstructured text data, such as categorizing emails, research papers, or customer reviews, leading to automated organization and efficient data retrieval. ML algorithms, both supervised and unsupervised, can be used to recognize patterns, detect anomalies, and make predictions within unstructured documents. By employing computer vision algorithms, organizations can automatically classify images, identify objects, and generate textual descriptions, revolutionizing fields like healthcare, security, and manufacturing.

Deriving value from unstructured data is a significant challenge, but leveraging Artificial Intelligence (AI), Machine Learning (ML), Natural Language Processing (NLP), and Computer Vision (CV) technologies can help unlock its potential. Here’s a high-level overview of how these technologies can be used:

Data Preprocessing: Before applying AI and ML algorithms, unstructured data needs to be processed and structured. This involves tasks like data cleaning, normalization, and transforming the data into a suitable format for analysis.

Natural Language Processing (NLP): NLP techniques can be used to classify unstructured text data into predefined categories or topics. This can enable automated categorization and organization of large amounts of textual information. Then by Named Entity Recognition (NER), algorithms can identify and extract entities like names, locations, organizations, and other relevant information from unstructured text. AI models then analyze text sentiment to determine whether it’s positive, negative, or neutral. This can be useful for understanding customer feedback, social media sentiment, or market trends. NLP techniques can also automatically generate summaries of large documents or text datasets, enabling quick extraction of key information.

Machine Learning (ML): ML algorithms can be trained on labeled data to recognize patterns and make predictions. For example, ML models can learn to classify images, identify objects, or recognize patterns in unstructured data. Through unsupervised learning, these algorithms can identify hidden patterns or clusters in unstructured data without any predefined labels. This can help in data exploration, segmentation, or anomaly detection. ML algorithms can also analyze user behavior, preferences, and unstructured data such as product reviews or browsing history to make personalized recommendations. Along with things, ML models can learn patterns from normal data and identify outliers or anomalies in unstructured data, which is particularly useful for fraud detection or cybersecurity.

Computer Vision (CV): CV techniques can classify and categorize images or videos based on their content, enabling automated analysis and organization of visual data. These algorithms can identify and locate specific objects within photos or videos. This can be useful in various applications, such as self-driving cars or surveillance systems. Such AI models can also generate textual descriptions or captions for images, enabling better understanding and indexing of visual data.

Use Cases

By combining these technologies, organizations can extract valuable insights, automate manual processes, improve decision-making, enhance customer experiences, and gain a competitive edge by making the most of unstructured data.These technologies can be used to analyze customer feedback from social media posts, reviews, or customer support interactions to understand the sentiment, identify emerging trends, and improve products or services. it can help organizations to automatically categorize customer queries or complaints to prioritize and route them to the appropriate departments for faster resolution. These algorithms can mine unstructured data from customer surveys or feedback forms to extract actionable insights and identify areas for improvement.

Analyzing unstructured data, such as transaction logs, emails, or support tickets, can help identify patterns indicative of fraudulent activities or cybersecurity threats. By applying NLP techniques it can be used to detect suspicious text patterns or anomalies in financial reports, insurance claims, or legal documents. By combining unstructured data sources like social media posts, news articles, and public records to assess reputation or compliance risks associated with individuals or organizations.

Using CV algorithms for facial recognition and object detection in surveillance videos to enhance security measures and identify potential threats or suspicious activities. Analyzing images from medical scans or remote sensing data can be used to assist in diagnosis, detect anomalies, or monitor environmental changes. ML and CV techniques can also be applied to monitor manufacturing processes, detect defects in products or equipment, and ensure quality control.

Extracting structured data from unstructured documents like invoices, contracts, or financial reports to automate data entry, streamline workflows, and improve operational efficiency. Automatically generating summaries or key insights from lengthy reports, research papers, or legal documents to aid in information retrieval and decision-making.

These use cases highlight the diverse applications of AI, ML, NLP, and CV in deriving value from unstructured data across various industries, including finance, healthcare, retail, manufacturing, and more. By harnessing the power of these technologies, organizations can unlock valuable insights, drive innovation, and gain a competitive edge in today’s data-driven landscape.

If you’re interested in exploring these technologies and their use cases further, don’t hesitate to reach out to us at open-innovator@quotients.com. We are here to assist you and provide additional information.

\

Categories

Categories
Applied Innovation

Artificial intelligence (AI)- the next stage in the transition from conventional to creative farming

Image Courtesy: Quotients
Categories
Applied Innovation

Artificial intelligence (AI)- the next stage in the transition from conventional to creative farming

Artificial intelligence (AI) has the ability to transform the way we think about agriculture by bringing` about numerous advantages and allowing farmers to produce more with less work.

With increasing urbanization with the world’s population and shifting consumption patterns, and rising disposable money, Farmering Industry is under a lot of strain to satisfy the rising demand and needs to find a method to boost output. There is a need to search for methods to lessen or at the very least control the risks faced by farmers. One of the most interesting possibilities is the application of artificial intelligence in agribusiness.

Artificial intelligence (AI) is the next stage in the transition from conventional to creative farming. Here we are discussing some applications of AI in agriculture:

Soil and Crop Monitoring

The amount and quality of the yield, as well as the health of the product, are directly influenced by the micro- and macronutrients in the soil.

In the past, personal sight and opinion were used to assess the health of the soil and the crops. However, this approach is neither precise nor prompt and in its place, UAVs can now be used to collect aerial picture data, which can then be fed into computer vision models for intelligent agricultural and soil condition tracking. This data can be analyzed and interpreted by AI much more quickly than by humans in order to monitor agricultural health, forecast yields accurately, and identify crop malnutrition.

Farmers typically have to collect soil samples from the ground and transport them to a facility for labor- and energy-intensive analysis. Instead, researchers chose to investigate whether they could teach a program to perform the same task using image data from a low-cost handheld camera.

The computer vision model was able to produce approximations of sand composition and SOM that were as accurate as pricey lab processing.

Therefore, not only can computer vision remove a significant portion of the labor-intensive, manual work involved in crop and soil track, it often does so more efficiently than people.

Monitoring Crop Maturation

To maximize output effectiveness, it is also crucial to watch the growth stages. To make changes for better agricultural health, it’s essential to comprehend crop development and how the climate interact.

Precision agriculture can benefit from AI’s assistance with labor-intensive processes like manual development stage monitoring. For producers, overserving and overestimating agricultural development and maturity is a difficult, labor-intensive task. But a lot of that labor is now being handled with ease and remarkable precision by AI.

The farmers no longer needed to make daily trips out into the fields to inspect their crops because computer vision models can more correctly spot development phases than human observation. Computer vision can determine when a crop is mature by using an algorithm that examined the hue of five distinct crop components, estimated the crop’s maturity, and then used this information.

Detecting Insect and Plant Diseases

Plant pest and disease monitoring can be mechanized using deep learning-based picture recognition technology. This works by creating models of plant health using picture categorization, detection, and segmentation techniques. This is accomplished by using pictures of rotten or diseased crops that had been labeled by botanists according to four main phases of intensity to training a Deep Convolutional Neural Network. The substitute for machine vision entails extensive, time-consuming human searching and review.

Livestock Monitoring

Farmers can keep an eye on their livestock in real time by using AI. Dairy farms can now separately watch the behavioral characteristics of their cattle thanks to artificial intelligence (AI) solutions like image classification with body condition scores, feeding habits, and face recognition. Additionally, farmers can keep track of the food and water consumption as well as the body temperature and behavior of their animals. These benefits of AI are the main reasons the farming industry is seeing a sharp rise in demand for it.

Conclusion

Technology has been employed in farmland for a very long time to increase productivity and lessen the amount of demanding manual work needed for farming. Since the advent of farming, humankind, and agriculture have evolved together, from better plows to drainage, vehicles to contemporary AI.

Computer vision’s expanding and more accessible supply could represent a major advancement in this area. Because of the significant changes in our climate, environment, and dietary requirements, AI has the potential to revolutionize 21st-century agriculture by boosting productivity in terms of time, labor, and resources while also enhancing environmental sustainability. By implementing real-time tracking to encourage improved product quality and health, it is also enhancing agriculture.

Categories

Categories
Applied Innovation Industry 4.0

How AI is impacting the textile industry

How AI is impacting the textile industry [Image Courtesy: Pixabay ]
Categories
Applied Innovation Industry 4.0

How AI is impacting the textile industry

The textile industry is looking for high-quality low-cost strategies to differentiate itself and take production ROI into consideration due to growing competition led by high labor expenses.

Problems with the manual approach

Most of the time, in the textile industry, inspecting yarn is done manually, which takes a lot of time and work. Operators stationed at various lines do the check, they have to pick up a variety of goods at random and examine them with their unaided eyes. They determine the fibers’ grade by visual inspection and separate them accordingly. Due to this manual approach wide variety of faults, including stains, deformation, knots, broken yarn, splitting, fuzzy edges, and incorrect color, missed inspections are also rather prevalent. Rule-based vision systems are prone to high rates of incorrect detections and require manual double-checking when errors are irregular or occur in large numbers. Yarn inspection requires a more dependable approach in order to increase labor productivity.

AI-powered solution

Using an AI-powered solution several kinds of yarn flaws may be detected and identified through picture analysis. The AI model can swiftly and precisely find faults to increase detection rates and manufacturing output while lightening the load on manual inspection. The technology can continue to refine the AI recognition process as the amount of accessible data grows and enable the speedy transfer of training results into multiple manufacturing lines.

This involves the use of an appearance tester that tests the samples of yarn taken from the lab. Vectors are defined and utilized as network inputs, and sample photos are taken and preprocessed. Then feed-forward neural networks are employed that had been trained using the back-propagation rule. Better outcomes may be achieved by using a multilayer neural network in conjunction with picture enhancement to estimate various yarn metrics. As a result, a modeling system may be effectively constructed.

Artificial intelligence (AI) can also be used in forecasting yarn properties and dye recipes. By analyzing past data on dye recipes and their outcomes and extrapolating this understanding to the characteristics of new dye recipes, AI may also be used to predict the quality of dye recipes. Machine learning algorithms or other types of AI techniques may be used to achieve this. To identify trends and forecast results, one method is to employ machine learning algorithms, which can be trained on vast datasets of yarns, textiles, and dye recipes. A machine learning model, for example, may be trained on a dataset of fibers with known attributes, such as strength and fineness, in order to predict the characteristics of new fibers. Similar to this, a machine learning model that has been trained on a dataset of well-known yarn attributes like tensile strength and elongation may be used to predict the properties of new yarns.

The advantages of implementing a machine vision system include an inspection accuracy of about 98 percent and each product may be inspected. Overall, the use of AI for yarn property prediction, fiber grading, and dye recipe prediction may assist manufacturers in improving the quality and effectiveness of their processes, resulting in significant cost savings and improvements in product performance.

Our innovators have developed solutions for computer vision for inspection and forecasting yarn properties and dye recipes. Some of these solutions have been successfully implemented at different levels. Please write to us at open-innovator@quotients.com to know more about these solutions.

Categories

Categories
Applied Innovation Industry 4.0

Visual Inspection System improving Production Process

Categories
Applied Innovation Industry 4.0

Visual Inspection System improving Production Process

Manufacturing operations attempt to provide the greatest level of quality at all stages of the manufacturing or assembly process. This requires quality checks that need visual confirmation to verify that the pieces are in the correct places, have the correct form, color, or texture, and are free of flaws such as scratches, pinholes, foreign particles, and so on. Because of the volume of inspections and product variation, as well as the fact that flaws can appear anywhere on the product and be of any size, automating these visual quality checks is extremely challenging.

A visual inspection enables the production process to be improved. The vision inspection system has sensors and cameras and relies upon computer vision technology. A visual inspection machine compares two objects in order to provide a response or result. For performing the comparison, vision inspection systems contain all of the information required to categorize all of the items included in the inspection.

To decide which elements will pass the comparison and which will fail, the visual inspection machine incorporates photographs of past tests that were deemed successful. The cases of ideal elements, the elements to be classed, and those that will pass the tests are included in the system, both visually and by information from another class. Those elements that are included through images serve as a guide to be able to compare all the others, as stated earlier this is the visual comparison.

Visual comparison occurs when one piece is placed next to another, observed from multiple perspectives, and able to produce some form of link between them. The qualities to be compared are observed from various perspectives, first to make them correspond in their orientation and allow for a more accurate comparison, or it may also be accomplished by re-creating the photos in 3D format, superimposing one over the other, and describing the differences.

We have solutions for object recognition, fault detection, and process control. Our visual defect and dimensional sorting solutions for a variety of items deliver high productivity with excellent product handling and better inspection efficiency. These solutions are designed with the customer’s needs in mind and are highly customizable to the user. It is combined with cutting-edge in-house software that is dependable and accurate, resulting in the best final product.

Some key features of the solution are:

– Real-Time Monitoring
– Resident Database SQL
– Remote Control and Setting
– Vision Software Programmable by User
– Statistical Reports easily accessible through UI
– Machine Vision Software
– Accuracy up to ±10 microns
– Support for multiple Camera units
– Run as a Turn-Key system
– Simple Setup and Adjustments
– Short Setup Time for multiple codes

To learn more about this product and how you can use it to grow your business, increase productivity, improve quality, and create a better working environment for your employees, please contact us at open-innovator@quotients.com.

Categories

Categories
Applied Innovation Industry 4.0

Intelligent Video Analytics Transforming Manufacturing, Hospitality, Transportation, Healthcare, Retail, etc

Image Courtesy: Twitter
Categories
Applied Innovation Industry 4.0

Intelligent Video Analytics Transforming Manufacturing, Hospitality, Transportation, Healthcare, Retail, etc

Video analytics, also referred to as video content analysis or intelligent video analytics, in recent times has been at the center of attention in both the industry and the academic world.

Advances in Deep Learning aiding Video Analytics

With advances in Deep Learning research and expanded availability of video data, video analytics now allows for the automation of tasks that were once possible by a human intervention only. This allows it to be used in a number of applications ranging from monitoring traffic jams and alerting in real-time, to analyzing customers’ flow in retail to maximize sales.

In Deep Learning, a subset of AI, a machine is exposed to volumes of tagged data allowing it to learn and recognize and identify the same information in new data sets imitating the way a human works. Deep Learning offers advantages like faster analytic output, improved processing performance, and increased object detection, accuracy, and classification.

Intelligent video analytics automatically recognizes temporal and spatial events in videos and performs real-time monitoring but it can also be used to analyze historical data to find insights. It can recognize objects, object attributes, movement patterns, or behavior related to the monitored environment are detected.

Some applications

Video analytics has the potential to be widely used in industries such as manufacturing, hospitality, education, retail, and others. We are discussing a few of them.

Healthcare

Integrating video analytics into legacy CCTV systems can transform cameras into much more proactive intelligence tools that can be used to ensure the safety of patients, staff, and visitors. Some of the most common problems like theft, infant abduction, and drug diversion can be detected and checked.

Mental healthcare is another area in which video analytics can be used to analyze facial expressions, and body posture to alert the hospital staff. It can also play a role in the at-home monitoring of older adults or people with health issues

Further, the data collected can be used to generate insights that can help to shorten wait times and achieve business goals by managing the staff according to patterns in the footfall of patients.

Transportation

Video analytics can be used in reducing accidents and traffic jams by dynamically adjusting traffic light control systems by monitoring traffic. By recognizing situations that may turn fatal in real-time, it can raise alerts, and even in the case of an accident, these systems can trigger an alarm to security and healthcare institutions to take action apart from that it can also serve as evidence in case of litigation.

Video analytics can also perform tasks like vehicle counting, speed cameras can detect traffic movements and license plate recognition can spots stolen vehicles or vehicles being used in a crime. It can also generate high-value statistics to assist in making infrastructure-related and other policy decisions.

Retail

The retail industry can use video analytics to generate insights and actionable information on customers’ behavior and buying patterns through their key characteristics like gender, age, duration and time of visit, walkways, etc. These algorithms can also be used to recognize previous customers and improve customer experience and provide personalized service. Video analytics can also play a role in developing anti-theft mechanisms by identifying shoplifters.

Manufacturing

Video analytics can improve productivity, reduce downtime and ensure staff health at the manufacturing facility by enhancing operations and management efficiency.

Smart cameras can be used to predict potential interruptions, evaluate specific bottlenecks and reduce downtime by generating alerts to take proactive action immediately. It can also optimize the number of employees in the production facility and improve overall productivity. Inventory management can also be enhanced by analytics as the warehouses can be monitored for their capacity. The use of machine vision can help in inspection and improve quality control.

Video analytics can also warn of situations that may pose threats to people, products, or machines by detecting movements and identifying conditions. Video analytics can provide round-the-clock security and alert commercial as well as residential buildings and prevent potential break-ins.

Video Analytics Approach

Video content analysis can be done in real-time or post-processing. Also, it can be centrally on servers that are generally located in the monitoring station or can be embedded in the cameras themselves, some times a hybrid approach is adopted.

There are startups that are working on Video Analytics and have successfully deployed their solutions across various sectors such as hospitality, retail, manufacturing, pharma, and food. To more about evolving use-cases and startups in different domains please write to us at Open-innovator@Quotients.com

Categories

Categories
Industry 4.0

AI-powered Computer Vision Revolutionizing Multiple Industries

Image Courtesy: Twitter
Categories
Industry 4.0

AI-powered Computer Vision Revolutionizing Multiple Industries

Inspections are critical for attaining manufacturing excellence. Inspection of processes and products determine the business success and customer confidence in the brand. Companies are increasingly relying on next-generation inspection solutions to improve quality control and deliver defect-free products. 

AI-powered Computer Vision

Digitization and machine vision based on AI algorithms can identify manufacturing anomalies much faster than human inspectors and improve quality and reduce costs. It employs computer vision, a field of artificial intelligence (AI) that enables computers and systems to obtain useful insight from digital images, videos, and other visual inputs — and take actions or make recommendations.

AI-backed computer vision is finding applications in industries ranging from manufacturing to automotive – and the market is expanding rapidly. It can perform the functions like inspection and identification in much less time. By the use of cameras, data, and algorithms, a system can be trained to inspect products or watch a production asset and analyze a very wide range of products or processes and detect invisible defects at a rate exceeding human capabilities.

Deep learning, an aspect of machine learning technology, trains machines by feeding a neural network with examples of labeled data. This is used to identify common patterns based on these examples and then convert it into a ‘math equation’ that mimics a human visual inspection classifies forthcoming information and performs tasks like differentiating parts, abnormalities, and texture.

Use Cases:

Automatic Counting: Computer Vision can be used for counting applications in industries where small parts are manufactured in large numbers like in metal parts, foods, pharmaceuticals, food, rubber pieces, wooden products, jewelry, etc.

Detect absence/presence: Computer vision can also detect the absence and presence of something such as date print, tags, brand logos, codes, stamps, etc, and automatically confirm the completeness of the product.

Sorting: Vision systems powered by AI algorithms can identify the right and defective product types by imaging them at high speeds. For example, separating pills in the pharma industry and segregating broken and damaged items in jewelry. This can be followed by sorting the identified items into chosen categories.

Surface Inspection: Computer Vision can identify surface anomalies for example scratches, dents, and pits accurately and at a high speed. Defects in some products like fabrics or automobile bodies are very small and undistinguishable, which can be detected only by monitoring the variation in intensity using deep learning algorithms.

Application:

Machine Vision is powering Industrial Automation. Using the latest 2D, 3D, and Artificial Intelligence solutions inspection systems are used across various industries like the pharmaceutical industry, automotive industry, printing and packaging industry, food and beverage industry, and textile industry. It offers huge benefits in eliminating human interventions and errors thus cutting down heavily on inspection cost and time.

There are startups working on this solution helping the above-mentioned as well as other sectors to greatly enhance their functioning through the acceptance and integration of new technologies into their existing systems. To know more about these and for collaboration and partnership opportunities please write to us at open-innovator@quotients.com

Categories

Categories
Innovator's Vista

How Smartphones are leading to Decentralization and Democratization of Diagnostics

Image Courtesy: NeoDocs
Categories
Innovator's Vista

How Smartphones are leading to Decentralization and Democratization of Diagnostics

India is witnessing an unprecedented surge in demand for E-healthcare platforms post coronavirus outbreak. Several health tech startups have come up that are addressing various customer pain points.

Diagnostics is the starting point in healthcare delivery and is fundamental to all the steps thereon. The diagnostics industry is also not untouched by what is referred to as the fourth industrial revolution. Developments in the fields such as biotechnology, genetics, and nanotechnology, are leading to an accelerated rate of change that generates new and very disruptive technologies. With advances in genomics, personalized medicines, and patient-centric approaches, AI-based data-driven tests have the potential to address many customer pain points. These factors are forcing the clinical laboratories to adjust and change their processes.

India’s diagnostic industry is one of the fastest-growing services in the country estimated at USD 9bn (around INR 675bn), still, the industry is highly fragmented and under-penetrated. There are 1.1 lakh medical laboratories in the country, whose test reports determine over 70 percent of medical decisions and of these, just about 1,039 are accredited and around 80-85 percent of these labs are unorganised. There is a rising need for this scenario to change and a more user-friendly, better, and transparent service/delivery is available to the user.

India will have 1 billion smartphone users by 2026 with rural areas driving the sale of internet-enabled phones, a Deloitte study suggests. The country had 1.2 billion mobile subscribers in 2021, of which about 750 million are smartphone users. Smartphone‐based diagnostics thus has the potential to lead to decentralization and democratization of clinical laboratory tests. It can allow the delivery of precise diagnostics in remote areas and limited resource settings practically possible.

Some startups are using standard smartphone cameras and a dipstick for urine tests and an app to identify different health indicators. All the patient needs to do is take a photo of the dipstick against a color card, and the app then gives an instant analysis.

NeoDocs, a Mumbai based Startup, is working on this use-case. The company is using technologies such as Lateral Flow Assay and Colorimetric Assay for this. Lateral flow assays (LFAs) are the technology behind low-cost, simple, rapid, and portable detection devices. It is a paper-based platform for the detection and quantification of analytes in complex mixtures, where the sample is placed on a test device and the results are displayed within 5–30 min. Colorimetric Assay involves the generation of color by chemical/biochemical reaction between target analyte and reagents. The intensity of the resulting color can be quantified using imaging tools and processing software.

The computer imaging technology is used to analyze color-coded slides with a dipstick — a plastic strip treated with chemicals that changes color when dipped in a sample. The color-coded slide is then analyzed by the app using artificial intelligence and machine learning to perform medical tests in the cloud. To account for the type of camera, the lighting conditions, and other variables the app uses machine learning. The results of the tests are then automatically sent for clinical follow-up.

NeoDocs has started with urine-based tests that can be used for issues starting from Urinary Tract Infection, Pregnancy, nutrient deficiency, Chronic Kidney Disease, Jaundice, etc. It plans to expand to blood and saliva-based tests in the near future.

Open Innovator is a platform and service to discover the use cases and best-suited tech solutions in any enterprise or domain and engage with innovators and founders. Connect with us and Follow us for more updates on evolving use cases in various domains.

Categories

Categories
Innovator's Vista

Autonomous Mobile Robots improving Healthcare

Ajay Vishnu, Founder, Jetbrain at Open Innovator Meet
Categories
Innovator's Vista

Autonomous Mobile Robots improving Healthcare

Robots are no longer automated stationary machines executing specific tasks but have evolved into sophisticated, mobile platforms solving various automation needs. An Autonomous Mobile Robot (AMR) is one such example that can independently move through any environment understanding and performing various tasks.

It is an innovative automation solution performing assigned tasks in varied environments like construction sites, warehouses, healthcare centres etc. Often compared with Automated Guided Vehicles (AGV) but unlike them AMRs do not operate on predefined paths or require operator sight. With more advanced software and hardware, AMRs are a step above and beyond AGVs.

AMRs have higher level of autonomy and capacity of moving around in shared environment. They rely on cameras, a set of sensors, artificial intelligence, machine learning, machine vision and navigation technique like collision avoidance for functioning like identifying, picking, moving objects from one location to another while avoiding obstacles.

They are finding application in warehouses for tasks like pick-and-place items, in e-commerce applications for moving carts and in automated construction. Another use of AMRs is in healthcare sector for automating distribution of medicine to the correct patients in hospitals.

Jetbrain Robotics, a Gurugram based startup, is working on providing end-to-end solutions for hospitals to provide enhanced patient care, using autonomous mobile robots (AMRs). The company envisions AMRs that would assist with nursing and managing internal logistics, enhancing the productivity of hospital staff.

The company has designed an AMR to carry hospital utilities to different destinations as required. Named AMRO, the robot has a compact body and protective outer layer. It can charge itself and work all day with inbuilt power capacity, claims the company.

AMRO carries hospital utilities to different destinations as required

Another product by Jetbrain is a UV disinfection robot with an Ultraviolet (UV) lamp emitter able to clear and disinfect environments from bacteria and viruses. Lytbot emits powerful broad-spectrum UV light, targeting high-touch surfaces while disinfecting the whole room. ‘Core 0’ is another such offering by the company that takes care of internal logistics and carrying items like surgical instruments, glucose packets, blood samples and medical necessities to required destinations.

Founder of Jetbrain Robotics, Ajay Vishnu, was present in the Open Innovator Meet recently held at NASSCOM CoE, Gurugram. He at the meet briefed about the vision of the company and plans for the near future.

Categories

Categories
Innovator's Vista

Computer Vision for Quality Inspection

Categories
Innovator's Vista

Computer Vision for Quality Inspection

In manufacturing sector whether it is clothing and textiles, petroleum, chemicals and plastics, electronics, computers and transportation, food production, metal manufacturing etc, poor production quality has a significant impact on efficiency. It results in additional operational and financial costs in form of lost time, wasted resources, scraps and decreased efficiency. Therefore, maintaining quality standards is of utmost importance in the field of manufacturing.

Typically product is visually inspected for defects which is a highly manual process. It is time consuming, prone to errors and dependent on operator’s experience and requires consistency which is not always possible.  Rule-based visual inspection machines which are programmed, are not flexible, and cannot adapt to product changes and can only detect a handful of defects at a time.

By using computer vision in manufacturing quality inspection can be automated during the production process.  There are several advantages of using computer vision like reduced cognitive load for operators, no programming required, and it adapts to product changes. It also makes inspections faster, more accurate, and efficient. There are today several manufacturing firms that are shifting towards using deep learning and computer vision for quality control and inspection tasks. 

Computer Vision:

Computer vision seeks to replicate and automate tasks that the human visual system can do. It is an interdisciplinary scientific field that aims to give computers the ability to extract a high-level understanding of the visual world from digital images and videos. It uses Artificial Intelligence (AI) and deep learning models to enable machines accurately identify, interpret, understand, classify and react to objects in form of visual data like digital images from cameras and videos.

More sophisticated artificial intelligence (AI)–based vision systems can enable more powerful visual inspection solutions. These solutions can handle complex applications with less engineering time compared to traditional machine vision solutions.

These systems are faster and simpler process that can carry out repetitive and monotonous tasks at a faster rate, which simplifies the work for humans. Computer vision systems better trained through all kinds of data generally commit zero mistakes resulting in faster delivery of high-quality products and services. With no room for faulty products and services, companies do not have to spend money on fixing their flawed processes.

By using a laser coupled with a 2D camera, object edges can be more easily located and “pseudo 3D” images can be produced. These 3D machine vision systems creates a full 3D profile of the object by stitching together the individual lines of image data that makes inspection of complex assemblies and sub-assemblies as well as individual components easier.

Categories

Categories
Innovator's Vista Mobility

AI driven solution to assess road network

Image Courtesy: Road Metrics
Categories
Innovator's Vista Mobility

AI driven solution to assess road network

Transportation sector is one of the crucial sectors for global economy and road infrastructure is the backbone of the transport network. Road networks have a direct and substantial impact on people’s lives, and thus it needs for regular maintenance and exhaustive monitoring.

The conventional methods for road condition assessment are labor intensive and often fail to meet the present requirements due to the vast area of road networks to inspected within a limited timeframe. Usage of sensors by automation of road damage detection using high-performance sensors is also taking place at some levels but these options are often costly and sometimes unaffordable.

In this context, using AI driven solutions to assess road network for an automated solution can help in easy and early identification of road defects. It involves usage of technologies like Computer Vision, DeepTech, and Machine Learning.

With the advent of Smartphones and Object Detection techniques in AI, these AI techniques are being used to perform road damage inspection. Computer vision AI software extracts road and street level data and identifies information such as road health, street informatory/warning signage and assets.

A Startup, Roadmetrics.AI, is working on this use case. Users need to simply mount smartphone with a customized data collection mobile app above vehicle’s dashboard which captures an image every 10 feet with the entire view of the road. The image data capture starts automatically with the trip and and stops at the end. A specific route can be selected with a targeted approach.

The image data then is sent into a dedicated system, where algorithms—trained with hundreds of thousands of image data points ensuring a high accuracy and constantly improving— can look for damage to the road surface image by image, classify various stages of road pavement damage. Damages are recorded and transferred again to the app. The privacy is also taken care as sensitive information is automatically removed according to GDPR compliance using customised algorithm.

This approach results in considerable savings in time and resources spent by customers in defect identification. Road defects such as cracks and potholes that are a major problem and result in billions of dollars in repair and maintenance can be accurately detected and updated databases of recorded structural damages can be maintained for further maintenance. Some of the key customers for such products are Governments, Road development authorities, Private parties operating in infrastructure sector.


Categories