The Dawn of Cobot Integration in Tomorrow's Factories
In this edition of Factory Trends, we delve into the world of humanoid integrations, a topic that's captivated our team and community. While not our usual focus, it is the topic that has captured our imagination and dominated discussions even more so after reading about the progress of Covariant, the OpenAI spinoff that aims to create an AI model enabling robots to learn tasks ...like humans?
Robotic Flexibility in the Manufacturing Industry
Visiting clients often reveals how robotic arms could streamline repetitive tasks, yet challenges such as multi-task training and assessing return on investment (ROI) remain.
At this point, the training of the robot and the flexibility it possesses is critical. Currently, some robots can be trained simply, with grips that mimic the nozzles and techniques of stable diffusion. These models explore precision in picking up objects; handling delicate objects is becoming increasingly simpler, and there are robots equipped with high-end innovative sensors like those developed by Jesse Opoku, PhD from Wootzano that can even pick fruits or tomatoes.
What are the costs, and the implementation times? Those are the challenges. Mainly when not manufacturing in line, in large volumes.
I believe this is one of the main challenges; in SMEs, it's rare for a workbench to be dedicated to a single task, so a robot cannot be used for a single product.
Now, without going to the latest technology from Covariant, how can the learning process of a robotic arm be simplified using computer vision, for example?
Simulated hands, with different grips, that record the movement and replicate it, have greatly simplified this task. These devices could significantly accelerate the deployment of robotics, which is one of the challenges we face here in the UK, for example. Data from the International Federation of Robotics indicates that robotic density in the UK was reported to be 101 units per 10,000 workers. This figure places the UK below the global average. International Federation of Robotics (ifr.org)
A Paradigm Shift in Robotics
The shift from viewing robots as single-task tools to versatile assets is underway. The technology is changing dramatically.
With computer vision, it's possible to combine AI, GenAI, and LLMs so that they can understand the context in which they are performing a task. This will allow robots to be more versatile, and the manufacturer can quickly visualize the return on investment.
In a quick chat, Aron Kisdi shares insights on the ROI challenge: ‘ROI is key. This is often forgotten or ignored by robotics companies. If there is no ROI to the customer there is no point in automation. However, in many cases, ROI is possible either by using simpler, cheaper robotics, or robots that have more bandwidth or can do more. In robotics, we need to identify the problems that can be solved by robotics and provide the largest return first, then move to other challenges. Often the ability to re-assign staff to more meaningful tasks can lead to big returns. Or make people more efficient. We need to focus on increase of efficiency and production. Humanoid robots are great, it is good news that people now taking these seriously and there are many interesting things to come. Personally, I think dog robots are pointless and we should focus effort on humanoids and dexterous handling. However, media consistently overestimates current capability. We are a minimum 3 years, more likely 5 years away from humanoid robots doing very simple tasks. It also needs scale, we need to think in 100s of robots. It does however mean that large companies starting to invest and learn how this works (like BMW) but for small companies, it is too early. ‘
Towards Robotic and AI Regulation
领英推荐
While social media is flooded with videos of robots performing simple and repetitive tasks, and organizations continue to push forward with funds for robotization, manufacturers still look at the investment with scepticism. Solutions like Dafo emerge as a middle ground to bridge this gap, as robots become increasingly flexible and the robotization of certain tasks becomes as clear as it is inevitably necessary when the issue of filling a job position becomes constant. Some reports indicate that 74,000 vacancies have not been filled in the UK manufacturing industry.
Meanwhile, regarding AI news, Europe has taken a significant step in regulating artificial intelligence. The Council and the European Parliament have reached a provisional agreement on the Artificial Intelligence Act, which is the first set of rules for AI in the world. This law aims to ensure that AI systems used in the EU are safe, respect fundamental rights, and adhere to EU values. These regulations would apply to any company operating in the EU, similar to the GDPR standards.
Celebrating Pi Day and One Month of Factory Trends
Finally, we love numbers and maths; Gabriel is very good at it. Then, we cannot let it pass that tomorrow is Pi Day, but there’s an even better reason to celebrate 3/14 because Factory Trends turns 1 month old :)? We extend our heartfelt gratitude to all our subscribers and the pioneering minds in manufacturing who have engaged with our content.
What did you think of today's newsletter? How do you see robotics and AI transforming your work? Share your comments with us – we're listening.
Thank you very much for reading us, until next time,?
Daniela, CEO & Founder
Gabriel Giani, Co-Founder AI Specialist and Code Captain
The Dafo.AI Team
+Plus
Chief Operating Officer at Wootzano
8 个月Great read and some positive insights here. Tha ks for the mention Daniela
Generalist | #DigitalForGood ?? | There is not a Planet B ??
8 个月Thank you for your contribution to this edition Aron Kisdi ??