AI & IoT integration for smart factories
November 30, 2023 12:58 pm
The manufacturing sector is witnessing a steady transition with the advent of Industry 4.0. The Internet of Things (IoT) and Artificial Intelligence (AI) are credited with introducing this shift. Factories are moving towards smart manufacturing, where machine learning, IoT, and AI technologies are applied to automate processes and evaluate data. Our industry experts in smart manufacturing shared the processes and challenges encountered while converting factories into smart factories.
The synergy between AI and IoT is crucial for providing valuable outputs to the manufacturing industry. More than just IoT or AI in isolation is required; effective integration is key. IoT optimises communication across the entire automated system. Once data reaches a decision point, AI comes into play, assessing defects and analytical parameters. The collaborative efforts of IoT and AI work hand in hand to ensure a comprehensive approach, ultimately contributing to the customer’s success.
In the context of Industry 4.0 implementation, a critical aspect is coordination and communication within a manufacturing company. This proves challenging as various departments, such as quality, IT, production, and maintenance, may initiate the process independently. Mr. Avil Kumar, Head of Software Services, Carl Zeiss India, adds, “The confusion arises regarding where the organisation should begin its Industry 4.0 journey. Currently, different departments focus on specific modules without a cohesive long-term and short-term strategy. The key is establishing clear organisational goals and approaching the transformation step by step, considering different levels or phases.”
Another consideration in adopting Industry 4.0 is how we measure results for our Indian customers. In India, the approach to budgeting solutions and assessing return on investment is typically straightforward – for instance, investing a certain amount in machinery may yield a predictable 10% increase in production. However, a more nuanced understanding is required when it comes to IoT. The challenge lies in quantifying the return on investment for small manufacturing units in India. It is essential to articulate and demonstrate how these units can realise tangible benefits daily as they transform.
Mr. Chandrashekar Bharathi, Managing Director, AceMicromatic Manufacturing Intelligence Technologies Pvt. Ltd notes that Ace Manufacturing was a pioneer in the Industry 4.0 landscape of India long before the term gained popularity. Over the past decade, they have facilitated over 20,000 installations of connected machines and factories, allowing even SMEs to integrate these technologies into their daily operations. While current discussions about AI in smart factories often centre on predicting quality and maintenance, there is a broader scope to consider. Real-time visibility is crucial for customers, extending beyond anomaly predictions to actionable insights within the first 30 minutes of production. The ability to correct issues promptly leads to optimal throughput for the remaining 7 hours of an 8-hour shift.
Another aspect involves the vision of smart, connected factories, where smart materials and interconnected machines harmonise production processes. Despite these advancements, companies must focus on their core activities to generate daily revenue.
Ground Level challenge
There is significant interest in adopting new technologies in the manufacturing sector, but practical challenges exist at the ground level, especially in Indian factories. Anish Pandari, Director of product and strategy at Indxo Al (Automation Industry Association), notes, “The complexity arises from the coexistence of machinery spanning from the 1940s to the latest 2023 models, with a cultural inclination towards repairing rather than scrapping older machines. Despite this interest, there are genuine challenges in implementation.”
One key challenge is the misconception that digital transformation is a project rather than a continuous process. Many view Industry 4.0 as a failure when they treat it as a one-time project and fail to see immediate results. Achieving complete digital transformation is a gradual journey, requiring setting relevant benchmarks and celebrating small victories along the way. It is mandatory to ensure that individuals understand this process-oriented approach instead of expecting instant, comprehensive outcomes.
The industry often grapples with outdated equipment that needs connectivity features such as RJ45 ports or 0.5 sensors. Kundan Das, Founder, 4.0smartindustry Advisory Services Pvt Ltd, shares that these machines were not designed to generate data for income. In the current scenario, factory floors deploy traditional, older machines and modern counterparts with updated features and extensive connectivity. Despite needing more direct connectivity, even the older machines provide fundamental parameters like vibration, temperature, and current. A strategic approach involves extrapolating these basic parameters into a broader set of 30 parameters through specific devices. Industries can effectively integrate old and new Operational Technology (OT) devices and machines by maximising the intelligence derived from these parameters.
The advent of 5G, coupled with the demand for low latency and high connectivity, underscores the need to modernise the connectivity layer and integrate diverse devices within a factory, from sensors to training departments. Each device contributes valuable intelligence, and combining these insights is critical for addressing industrial challenges, including predictive maintenance and operational efficiency. The success of Industry 4.0 adoption is contingent on improving Key Performance Indicators (KPIs), as organisations are unlikely to embrace this transformative approach without tangible enhancements.
Before embarking on the journey of Industry 4.0 adoption, it is crucial to assess the current state of affairs. This assessment is vital as enterprises operate within budget constraints, and any allocation of funds must be justified. The initial step involves a meticulous audit of the shop floor, determining whether the organisation is at Industry 1.0 or 2.0. This audit follows the Smartness Index defined by the World Economic Forum. The Smartness Index involves analysing various touchpoints, including skill set, technology architecture, and productivity parameters. Through this analysis, the organisation can identify its current position and chart a course for the future. This planning involves understanding the required investment and projecting the anticipated improvements in Key Performance Indicators (KPIs). It is essential to recognise that every aspect is interconnected, and such strategic decisions are scrutinised closely by top management.
Your future advertising space? Our media data
Optimising supply chain management
There are two significant aspects to consider—one from the supply chain perspective and the other related to quality control on the shop floor. Due to heavy reliance on vendors, the supply chain plays a critical role in various sectors. Managing assessments manually with numerous vendors and parts is practically impossible. Introducing AI is essential to address such complexities. Recalling a specific case, Mr Pandari says, “We implemented an AI solution for real-time testing of every incoming product, reducing turnaround time significantly. This included in-line testing, which was previously challenging without destructive methods like dye painting or X-rays. The AI solution, utilising Sonic analysis, proved effective in resolving the issue. As supply chains become more intricate, the role of AI in various aspects, especially quality control, becomes increasingly crucial.”
The emphasis on quality is profound, allowing for a thorough examination of parameters. The intense focus on quality is a distinguishing factor for top brands like Toyota and Apple. Achieving such precision and reliability in output is only feasible with the integration of AI. The future demands a level of output and turnaround time that can only be achieved through the augmentation of processes with AI.
Mr. Bharathi considers that while evaluating the cost of automating or digitising a factory, the potential return on investment (ROI) and the specific goals a company aims to achieve are to be considered. Even if a company spends just ₹1 and does not see any tangible results, that extra rupee spent is considered an excess. To determine the cost, companies must assess their objectives—acquiring data from production processes, integrating metrology, or focusing on ROI as a pivot. Viewing ROI from the perspective of increasing Overall Equipment Efficiency (OE) reveals that a well-executed strategy can yield returns in weeks or months. This trend holds true across various industries, irrespective of company size or location. Timely access to relevant data with the right analytical approach significantly impacts productivity. While the exact cost varies based on specific goals, a benchmark range for customers could be tens of thousands to several lakhs of rupees per equipment.
Mr Avil notes that the initiation of upskilling efforts should commence at the operator level and extend throughout the organisation. The objective is to empower operators to a point where they are confident in making decisions. An illustrative example involves operators determining both the invoice quantity and value. Currently, we are progressing to a phase where operators will be responsible for deciding the entire dispatch quantity. This shift entails operators possessing knowledge about the machines, quality, and other aspects, with IoT stepping in to handle subsequent stages. The training process encompasses understanding machine operations, quality assessment, and proficiency in utilising various tools, gauges, and technologies. These parameters are essential considerations as we transition to and integrate new technologies.
Kundan Das, Founder, 4.0smartindustry Advisory Services Pvt Ltd.
“Maximising intelligence from fundamental parameters, integrating old and new OT devices, and adhering to Industry 4.0 standards are essential for transforming industrial operations and achieving tangible improvements in Key Performance Indicators.”
Mr. Avil Kumar, Head of Software Services, Carl Zeiss India.
“The symbiotic relationship between AI and IoT optimises communication and analytical assessment, which is pivotal for the success of Industry 4.0.”
Mr. Chandrashekar Bharathi, Managing Director, AceMicromatic Manufacturing Intelligence Technologies Pvt. Ltd.
“In Industry 4.0, real-time visibility is crucial for actionable insights, going beyond anomaly predictions to optimise production.“
Anish Pandari, Director, Product and Strategy, Indxo Al (Automation Industry Association)
“Digital transformation in manufacturing is a continuous process, requiring a process-oriented approach. Integrating AI, especially in quality control and supply chain management, is essential for addressing the complexities of diverse machinery, multiple generations of controllers, and intricate supply chains in the manufacturing sector.”