AI and Machine Learning

Decode hybrid AI system potential

New artificial intelligence (AI) software will contribute to the creation of more competitive sensor systems. Using ambient intelligence and a mix of AI tools will advance AI technology effectiveness. Hybrid AI can help with automation, manufacturing and robotics.
By Ashish Khushu October 5, 2019
Courtesy: L&T Technology Services Ltd.

Artificial intelligence (AI) platforms will trigger disruptive innovations across enterprises, from manufacturing to consumers. It is imperative to look into the widespread and complex issues AI can successfully tackle. In addition to capabilities in today’s software, AI is an amalgamation of several concepts, which can align and produce greater results.

In the industrial sector, AI has evolved over the years to produce a number of powerful tools, including knowledge-based systems, fuzzy logic, automatic learning, neural networks, ambient intelligence, and genetic algorithms. Applications of these tools in sensor systems have become more widespread due to the power and affordability of present-day computing systems.

Industrial AI tools include knowledge-based systems, fuzzy logic, automatic learning, neural networks, ambient intelligence, and genetic algorithms. Courtesy: L&T Technology Services Ltd.

Industrial AI tools include knowledge-based systems, fuzzy logic, automatic learning, neural networks, ambient intelligence, and genetic algorithms. Courtesy: L&T Technology Services Ltd.

Many new sensor systems applications may emerge as a result. Hybrid tools, which combine the strengths of two or more tools, may gain a greater role. Additional technological developments in AI will impact sensor systems include data mining, multi-agent systems, and distributed self-organizing systems. Appropriate deployment of new AI tools will contribute to more competitive sensor systems, known as hybrid AI systems.

Understand hybrid AI systems

The tools and methods driving hybrid AI have minimal computation complexity and can be implemented on small assembly lines, single robots, or systems with low-capability microcontrollers. These approaches use ambient intelligence and mix different AI tools to use the best of each technology, encompassing a more advanced framework when compared to traditional AI mechanisms.

As the name suggests, the purpose of a hybrid AI system is combining the desirable elements of different AI techniques into one system. Each method of implementing AI has strengths and weaknesses. Combining different methods can produce hybrid techniques with more strengths and fewer weaknesses. An example is the neuro-fuzzy system, which combines the uncertain handling of fuzzy systems with the learning strength of artificial neural networks.

AI system applications

Hybrid AI can help with automation, manufacturing and robotics, including weld programming. An existing system consists of two software systems working in series to construct viable robot programs to boost efficiency. The first system, the computer-aided design (CAD) model interpreter, accepts a CAD model and determines the welds required. This data is fed to the program generator, which re-orientates the weld requirements in line with the actual real-world orientation of the panel. The program generator sends any programs sequentially to the robot (normally one program per weld line). Additional software systems could be incorporated into the existing system at the point where the robot programs are sent to the robot system. At that point, the communication method is standard transmission control protocol/internet protocol (TCP/IP), and any programs sent can be viewed as text files.

Mixing sensor, logic systems

Researchers are mixing sensor systems and some powerful new technologies, with better results over time. Results include less use of energy, space, and time, along with more output for less cost. Machines read-in data from real objects and lay-down successive layers to build up an object model from a series of cross sections. AI reduces costs and time in most applications.

AI can increase effective communication, reduce mistakes, minimize errors, and extend sensor life.

Over the past 40 years, AI has produced a number of powerful tools, including those reviewed here: knowledge-based systems, fuzzy logic, automatic learning, neural networks, ambient intelligence and genetic algorithms. Applications of these tools in sensor systems have become more widespread due to the power and affordability of present-day computers.

More effective generation of robotic weld paths can result from artificial intelligence applications. Courtesy: L&T Technology Services Ltd.

More effective generation of robotic weld paths can result from artificial intelligence applications. Courtesy: L&T Technology Services Ltd.

Many new sensor system applications may emerge, and greater use may be made of hybrid tools that combine the strengths of two or more of the tools mentioned. The appropriate deployment of new AI tools will help create more competitive sensor systems. Ambient intelligence and a mix of AI tools applies the best of each technology.

Hybrid intelligent management systems

Over the past decade, industries have explored various opportunities to make the shift towards developing and applying hybrid intelligent management systems across various operations capable of using multiple AI techniques. It may take another decade for engineers to recognize the benefits due to a lack of familiarity and the technical barriers associated with using these tools, but this field of study is expanding.

Hybrid AI systems can deliver long-term sustainable business benefits throughout the industrial value chain. Company leadership needs to leverage available best-in-class solutions to transform legacy systems into modern-day models.

Ashish Khushu is chief technology officer, L&T Technology Services Ltd. Edited by Mark T. Hoske, content manager, Control Engineering, CFE Media, mhoske@cfemedia.com.

MORE ANSWERS

KEYWORDS: Artificial intelligence, AI for sensor systems

Hybrid AI creates value more quickly.

Automation, robotics, weld programming are among AI applications.

Computing power has helped AI advance.

CONSIDER THIS

What AI applications do you need to reconsider?


Ashish Khushu
Author Bio: Ashish Khushu is chief technology officer, L&T Technology Services Ltd.