Category robotics

Researchers have used artificial intelligence tools to accelerate the understanding of the risk of specific cardiac arrhythmias when various parts of the heart are exposed to different thresholds of radiation as part of a treatment plan for lung cancer.

A new adaptive 3D printing system can identify the positions of randomly distributed organisms and safely move them to specific locations for assembly.

On a research cruise around Hawaii in 2018, Yuening Zhang SM '19, Ph.D. '24 saw how difficult it was to keep a tight ship. The careful coordination required to map underwater terrain could sometimes lead to a stressful environment for team members, who might have different understandings of which tasks must be completed in spontaneously changing conditions.

It’s no exaggeration that nearly every company is exploring generative AI. 90% of organizations report starting their genAI journey, meaning they’re prioritizing AI programs, scoping use cases, and/or experimenting with their first models. Despite this excitement and investment, however, few businesses have anything to show for their AI efforts, with just 13% report having successfully moved genAI models into production. 

This inertia is justifiably causing many organizations to question their approach, particularly as budgets are crunched. Overcoming these genAI challenges in an efficient, results-driven manner demands a flexible infrastructure that can handle the demands of the entire AI lifecycle. 

Challenges Moving Generative AI into Production 

The challenges limiting AI impact are diverse, but can be broadly broken down into four categories: 

• Technical skills: Organizations lack the tactical execution skills and knowledge to bring Gen AI applications to production, including the skills needed to build the data infrastructure to feed models, the IT skills to efficiently deploy models, and the skills needed to monitor models over time.
• Culture: Organizations have failed to adopt the mindset, processes, and tools necessary to align stakeholders and deliver real-world value, often resulting in a lack of definitive use cases or unclear goals. 
• Confidence: Organizations need a way to safely build, operate, and govern their AI solutions, and have confidence in the results. Otherwise they risk deploying high-risk models to production, or never escaping the proof-of-concept phase of maturity. 
• Infrastructure: Organizations need a way to smooth the process of standing up their AI stack from procurement to production without creating disjointed and inefficient workflows, taking on too much technical debt, or overspending. 

Each of these issues can stymie AI projects and waste valuable resources. But with the right genAI stack and enterprise AI platform, companies can confidently build, operate, and govern generative AI models.  

Building GenAI Infrastructure with an Enterprise AI Platform

Successfully delivering generative AI models demands infrastructure with the critical capabilities needed to manage the entire AI lifecycle. 

• Build: Building models is all about data; aggregating, transforming, and analyzing it. An enterprise AI platform should allow teams to create AI-ready datasets (ideally from dirty data for true simplicity), augment as necessary, and uncover meaningful insights so models are high-performing. 
• Operate: Operating models means putting models into production, integrating AI use cases into business processes, and gathering results. The best enterprise AI platforms allow  
• Govern:

An enterprise AI platform solves a number of workflow and cost inefficiencies by unifying these capabilities into one solution. Teams have fewer tools to learn, there are fewer security concerns, and it’s easier to manage costs. 

Harnessing Google Cloud and the DataRobot AI Platform for GenAI Success

Google Cloud provides a powerful foundation for AI with their cloud infrastructure, data processing tools, and industry-specific models:

• Google Cloud provides simplicity, scale, and intelligence to help companies build the foundation for their AI stack.
• BigQuery helps organizations easily take advantage of their existing data and uncover new insights. 
• Data Fusion, and Pub/Sub enable teams to to easily bring in their data and make it ready for AI, maximizing the value of their data.
• Vertex AI provides the core framework for building models and Google Model Garden provides 150+ models for any industry-specific use case.

These tools are a valuable starting point for building and scaling an AI program that produces real results. DataRobot supercharges this foundation by giving teams an end-to-end enterprise AI platform that unifies all data sources and all business apps, while also providing the essential capabilities needed to build, operate, and govern the entire AI landscape

• Build: BigQuery data – and data from other sources – can be brought into DataRobot and used to create RAG workflows that, when combined with models from Google Model Garden, can create complete genAI blueprints for any use case. These can be staged in the DataRobot LLM Playground and different combinations can be tested against one another, ensuring that teams launch the highest performing AI solutions possible. DataRobot also provides templates and AI accelerators that help companies connect to any data source and fasttrack their AI initiatives,
• Operate: DataRobot Console can be used to monitor any AI app, whether it’s an AI powered app within Looker, Appsheet, or in a completely custom app. Teams can centralize and monitor critical KPIs for each of their predictive and generative models in production, making it easy to ensure that every deployment is performing as intended and remains accurate over time.
• Govern: DataRobot provides the observability and governance to ensure the entire organization has trust in their AI process, and in model results. Teams can create robust compliance documentation, control user permissions and project sharing, and ensure that their models are completely tested and wrapped in robust risk mitigation tools before they’re deployed. The result is complete governance of every model, even as regulations change.  

With over a decade of enterprise AI experience, DataRobot is the orchestration layer that transforms the foundation laid by Google Cloud into a complete AI pipeline. Teams can accelerate the deployment of AI apps into Looker, Data Studio, and AppSheet, or enable teams to confidently create customized genAI applications. 

Common GenAI Use Cases Across Industries

DataRobot also enables companies to combine generative AI with predictive AI for truly customized AI applications. For example, a team could build a dashboard using predAI, then summarize those results with genAI for streamlined reporting. Elite AI teams are already seeing results from these powerful capabilities across industries. 

A chart showing real-world examples of genAI applications for banking, healthcare, retail, insurance, and manufacturing.

Google gives businesses the building blocks for harnessing the data they already have, then DataRobot gives teams the tools to overcome common genAI challenges to deliver actual AI solutions to their customers. Whether starting from scratch or an AI accelerator, the 13% of organizations already seeing value from genAI are proof that the right enterprise AI platform can make a significant impact on the business. 

Starting the GenAI Journey

90% of companies are on their genAI journey, and regardless of where they might be in the process of realizing value from AI, they all are experiencing similar hurdles. When an organization is struggling with skills gaps, a lack of clear goals and processes, low confidence in their genAI models, or costly, sprawling infrastructure, Google Cloud and DataRobot give companies a clear path to predictive and generative AI success. 

If your company is already a Google Cloud customer, you can start using DataRobot through the Google Cloud Marketplace. Schedule a customized demo to see how quickly you can begin building genAI applications that succeed. 

The post Solving GenAI Challenges with Google Cloud and DataRobot appeared first on DataRobot.

VFDs are the backbone of modern and legacy industrial systems, managing tasks from powering fans and compressors to enabling intricate robotic operations. For more complex industrial operations, the need for precise coordination between drives escalates.

Over the past few decades, electronics engineers have developed increasingly flexible, versatile and highly performing devices for a wide range of real-world applications. Some of their efforts have been aimed at creating smart and sensing textiles, which could be used to fabricate stretchy robotic systems, medical devices and wearable technologies.

A collaborative approach to automation offers solutions where older methods have fallen short. Manufacturers who have struggled to achieve desired results from conventional robotics should consider how cobots can help them strike an ideal balance.

RobotWritersAI.com is playing hooky.

We’ll be back Sept. 2, 2024 with fresh news and analysis on the latest in AI-generated writing.

The post Gone Fishin’ appeared first on Robot Writers AI.

In recent years, roboticists worldwide have designed various robotic grippers that can pick up and manipulate different types of objects. The grippers that are most effective in tackling real-world manual tasks, particularly complex object manipulation tasks, are often those inspired by human hands.

A zinc-air microbattery could enable the deployment of cell-sized, autonomous robots for drug delivery within in the human body, as well as other applications such as locating leaks in gas pipelines.

Researchers have made groundbreaking advancements in the field of soft robotics by developing film-balloon (FiBa) soft robots. These innovative robots, designed by a team led by Dr. Terry Ching and corresponding author Professor Michinao Hashimoto, introduce a novel fabrication approach that enables lightweight, untethered operation with advanced biomimetic locomotion capabilities.

The great strength of Ensenso B lies in the very precise detection of objects at close range. It provides a wide field of view and an impressively high depth of field.

A tiny battery designed by MIT engineers could enable the deployment of cell-sized, autonomous robots for drug delivery within in the human body, as well as other applications such as locating leaks in gas pipelines.

A new algorithm may make robots safer by making them more aware of human inattentiveness. In computerized simulations of packaging and assembly lines where humans and robots work together, the algorithm developed to account for human carelessness improved safety by about a maximum of 80% and efficiency by about a maximum of 38% compared to existing methods.

A new algorithm may make robots safer by making them more aware of human inattentiveness. In computerized simulations of packaging and assembly lines where humans and robots work together, the algorithm developed to account for human carelessness improved safety by about a maximum of 80% and efficiency by about a maximum of 38% compared to existing methods.