Despite the infrastructure, hardware, and skillset challenges, India is navigating its way to automation by adopting Connected Intelligent Machines
The person who invented the wheel was, definitely, a genius. But what about the person who invented those three more wheels?
Machines have been around us for so long now that we cannot even imagine how our ancestors ever managed without them. But just a few decades from now, our progeny would wonder the same thing: how did we do without CIMs – Connected Intelligent Machines – for so long?
Security and control issues are key concerns, especially since the smarter a machine gets, the easier it can become for attackers to exploit a vulnerability.
The three new wheels here are – connected, intelligent and easy to use, and very soon we would see their various avatars throbbing inside our factories, mines, oilfields, highways and whatnot.
WHAT ARE THEY?
We already had machines. So, what makes this new species different from what they have been doing so far? The definition, interestingly, changes as per the context but the essence is similar.
In KPMG’s assessment, when we think of CIMs, we are thinking of an ecosystem of machines and plants that can talk to each other via IoT or any intelligent technology. Earlier, Individual machines could be sitting in India and Germany with no connection but now connectedness is a game-changer.
Human mediation is required between the output of a machine and the input of another, especially when the output is not deterministic.
Ask Sridhar Gopalakrishnan – Senior Vice President, Hexaware Technologies and he says that a CIM is an autonomous device that is part of a large network of such devices. It is intelligent in the sense that a CIM will make decisions on its own, respond to incentives embedded in the network, maintain trustworthiness and learn from other devices.
“Machines today are specific operators with clearly defined process steps to follow. Human mediation is required between the output of a machine and the input of another, especially when the output is not deterministic. With CIMs, this mediation will be reduced substantially. CIMs will also be context-aware and take personalised prescriptive actions.”
“CIMs will dynamically create personalised consumer experiences based on user contexts and interactions with other CIMs in the value chain.”
Sanjay Lodha, Chairman and Managing Director, Netweb Technologies explains this new paradigm from a data perspective. “The deluge of data is everywhere. Needless to say, we will need many intelligent machines to make sense of the continuous flow of incoming data. In recent years, the popularity of CIM has been gaining traction, with the concept of IoT becoming a reality and no longer just a theoretical construct. As per a report by Ericsson, CIMs are expected to be a crucial part of life by 2030, with consumers’ predictions.”
DOES 5G MATTER?
In the reckoning of Sam Fenwick, Principal Analyst, Opensignal, the importance of 5G and connectivity to a CIM depends on several interrelated factors including the machine’s function and location, where its intelligence resides and whether the flow of data between it and the user (or other machines) needs to take place in real-time, and how long the machine can function in the absence of connectivity.
“If the machine just requires connectivity for infrequent software updates and the transfer of information on a non-real-time basis, then the consequences of an hour-long interruption in connectivity are very different to those experienced when a drone is being used to provide real-time aerial footage of an ongoing incident.”
For telecom companies too, this change is a big shift as they are the key enablers through the Industrial Internet of Things (IIoT) and 5G, a KPMG report points out. 5G, Artificial Intelligence (AI) and IIoT are adjacent technologies that are big enablers for connected intelligent machines.
Fenwick argues how the degree to which the users of a machine can expect it to nearly always be connected strongly influences the importance of connectivity to its function. “A machine that will always stay within the confines of its owner’s premises and can be supported by a hardened private network — one that can be engineered to eliminate all coverage black spots — is at one extreme of this continuum. In such a case, it can be designed and used under the assumption that connectivity will always be present, potentially allowing much of its intelligence to be present in the cloud — thereby reducing the machine’s cost and complexity. However, if the machine’s function requires ultra-reliable and low latency communications, then its intelligence has to sit closer to the network edge.”
CIM frequently need to transmit a vast amount of data quickly, Lodha explains. “The prominent features of 5G, such as high data rates, low latency, network slicing and increased connectivity are ideal to ensure that CIM survives and thrives in the long run. IIoT plays a crucial role in enhancing productivity, supporting predictive maintenance, ensuring safety and supply chain and inventory management.”
“Prominent features of 5G, such as high data rates, low latency, network slicing and increased connectivity are ideal to ensure that CIM survives and thrives.”
“One of the key things to understand about 5G in this context is that it is still developing as a technology. For example, work is underway on 3 GPP Release 18 — the first Release to be branded as 5G Advanced. In addition, while the first 3GPP Release to include the standards for 5G was Release 15, the bulk of the major work on supporting edge computing with 5G didn’t start until 3GPP Release 17.” Fenwick reasons.
In addition, the bulk of initial 5G deployments worldwide focused on using 5G new radio technology in combination with 4G core networks, referred to as 5G non-standalone access, as opposed to 5G standalone access — where 5G new radio is used together with a 5G core network. Upgrading to the latter is complex and difficult, but enables new capabilities such as network slicing.
CHANGING THE GAME
When we comb through the 2022 KPMG Whitepaper, Connected Machines: A Game Changer in Industrial Manufacturing, we see that these machines bring several advantages to different entities.
Cost savings for manufacturers. Remote maintenance for dealers and service providers. Reduced downtime and service costs for end customers. For instance, Siemens Electronics Works, Germany saw a 140% jump in their factory output between 2017 and 2021. Similarly, Schneider Electric, Indonesia saw a 70% improvement in supplier service rate.
Their strength lies in what they bring to enterprises, the KPMG report reminds us. From reducing downtime to boosting efficiency to even enhancing end-customer experience, they bring a lot of advantages. Even unexpected ones: the need for less maintenance staff and hours, since everything is automated, or adding a new revenue pipeline through servitisation. Now someone who sells a compressor to a company can also offer proactive repair, thanks to real-time updates on the equipment’s health.
It is not just heavy-duty industries where they find relevance. Even verticals like healthcare, FMCG, food and retail can tap them for various outcomes. Manoj Gupta, associate vice-president of IT at Restaurant Brands Asia (formerly known as Burger King India) adds how CIMs can streamline order processing, reducing waiting time and enhancing the accuracy of orders. This is particularly valuable in high-traffic quick-service restaurants.
“Intelligent machines can handle routine customer interactions, such as taking orders and processing payments, freeing up human staff to focus on more complex tasks and providing better customer service. These Machines can collect and analyse data on customer preferences, ordering patterns, and operational efficiency, enabling data-driven decision-making for menu optimisation and business strategy. Also, they can monitor and predict equipment maintenance needs, reducing the risk of equipment failures that could disrupt service.”
“CIMs can streamline order processing, reducing waiting time and enhancing the accuracy of orders, particularly in high-traffic quick-service restaurants.”
Even leasing companies can tap this new form of intelligence. Case in point, how Hexaware enabled improved tracking of vehicles for one of the largest leasing companies in Europe finding difficulties in monitoring the leased vehicle’s location, thereby impacting the bottom line due to the high cost of recovery.
Sensors were attached to the vehicles to capture their location using a telematics device. Data points about harsh braking, engine temperature, idle time and distance travelled were captured. The idle time data was integrated with the Idle fuel flow factor to calculate the fuel savings. Information and details about every breach were recorded. The details helped reduce the idle time for leasing the vehicles. According to the company, the initiative also helped save fuel costs by 2–8% and maximise equipment uptime and lifespan.
CIMs are the next stage of evolution in increasing productivity, emphasises Gopalakrishnan. “In manufacturing, CIMs can run completely autonomous production lines and supply chains, propagating the response to market movements across the value chain. In services industries, CIMs will dynamically create personalised consumer experiences based on user contexts and interactions with other CIMs in the value chain.”
ARE THEY EASY TO START?
Not really. There are a host of aspects to consider before enterprises can think of putting their feet up and leaving everything in the hands of these smart pixies around. With more machines talking to each other with sensors and software sprawled everywhere, security becomes a big chink in the armour now. Recent data by SonicWall shows a 133% surge in ransomware along with a 311% jump in IoT attacks in India alone.
KPMG’s analysts also point out security and control issues as an important concern, especially since the smarter a machine gets, the easier it can become for attackers to exploit a vulnerability. Plus, chances are thin that an enterprise would go this route via greenfield investments. Most just choose to enhance existing infrastructure through the addition of sensors and connectivity.
Also, integration issues are a big fence to cross for anyone thinking of embracing this huge shift. “Before raising any process to buy any such machines, we find out the possibilities of integration with the system. So as soon as we receive such machines we connect into the system for its optimal use,” shares Bhoopendra Solanki, Chief Information Officer, Sakra World Hospital.
“Integration issues are a big fence to cross for anyone thinking of embracing CIM and one needs to explore this before buying any such machines.”
Then there is the gap in interoperability. This area requires the creation of common ontologies and data formats, Gopalakrishnan contends. “Without standardisation, we should be concerned about how CIMs interpret information. For CIMs to be effective, interpretations must be commonly understood and specifically implemented by each machine. Its actions will influence other machines in the network. So, it is important to ensure that the resulting network complexity does not become unmanageable. Detecting path dependencies, simulating outcomes and taking early corrective steps will be crucial.”
Not to forget, the industry would need new skills to steer these new wheels too. Sanjoy Paul, Program Director – Technology, Hero Vired. “The adoption of connected intelligent machines across industries, especially in manufacturing, would necessitate addressing several skill gaps and require upskilling in key areas. Technical expertise in Full Stack Development (FSD) would be essential. Full Stack Developers possess the knowledge and skills required to work on both the front-end and back-end of applications and systems. This proficiency is vital for building and maintaining the software components of connected intelligent machines.”
He also adds how a strong understanding of DevOps practices would be crucial to ensure operational efficiency. Proficiency in system integration and management is key.
“Full Stack Developers and DevOps professionals play a pivotal role in integrating new technologies with existing systems and ensuring their seamless management. This involves ensuring that all components work cohesively and efficiently, minimising disruptions in operations.” Lodha avers: “While the talent pool is vast, there needs to be more advanced, specialised skills for the next generation of connected technologies.”
Lodha also highlights areas like visibility and privacy. “The visibility issue arises when the CIM are not transparent or accountable for their actions and decisions. Some CIMs can use complex and opaque algorithms that need help understanding or interpreting. This can make verifying, validating, or debugging these machines and trusting their outcome or behaviours difficult.”
Since the CIM gathers data, there are pertinent questions regarding who is accessing the data, how it is being used, and if it is sold or shared with third parties, he warns. “Moreover, there is also the possibility of misuse in surveillance and monitoring, resulting in invasive tracking of individuals without their knowledge or consent.” He also observes how the existing law does not cover the intricacies and challenges connected intelligent machines pose. “New regulations are needed to ensure the safe and just use of technology.”
However, current signs of adoption hint that interest and investment, are pouring strongly into this lane. KPMG’s analysts have noted how companies are already, visibly, implementing such machines. Paul echoes the growing interest. “We have observed a notable surge in demand for skills in FSD, DevOps, and cybersecurity as industries increasingly embrace connected intelligent machines.”
“Expertise in front-end and back-end applications and systems is vital for building and maintaining software components for connected intelligent machines.”
Gupta also lets on the path laid out ahead. “Regarding our own QSR operations, we are actively considering and, in some cases, already using CIMs. We recognise their potential to improve efficiency, reduce costs, and enhance the overall customer experience. However, the specific implementation and extent of their use may vary based on the type of QSR, its size, and the technological infrastructure in place. The pace of adoption will depend on factors like the availability of suitable technology solutions, cost considerations, and alignment with our business goals.”
India is getting into top gear here too.
As digitisation deepens across industries, CIM opens tremendous opportunities for creating new AI models and, ensuring interoperability across systems, data standardization, security and network optimisation, portends Gopalakrishnan. “The government’s infrastructure, ‘Make in India’ and ‘Digital India’ investments augur well for the future of CIM tech development in the country.”
Lodha seconds that. “India has a booming hardware and electronic manufacturing industry duly supported by government initiatives. Many global companies have already set up assembly plants in the country. India is already home to several global and domestic OEMs in the automobile, electronics, and industrial machinery sectors. These OEMs adopt connected intelligent machines to improve productivity, efficiency, quality, and safety.”
He also points out that Hindustan Unilever Limited has launched the ‘Reimagine HUL Program’, which aims to transform its business from a linear FMCG value chain to an intelligent enterprise built on a connected consumer, customer, and operations ecosystem. “However, India will need more OEMs in other sectors, such as healthcare, education, and entertainment, to create more demand and innovation for connected intelligent machines,” Lodha says.
We would also need some more work to be done on the chip level. As Lodha notes: “Companies such as Micron Technology, Vedanta, and Foxconn are actively exploring the possibilities of setting up large semiconductor chip plants in India. However, as of now, India has a limited capacity for designing and manufacturing chips for CIMs. Most of the chips used in India are imported from countries like China, South Korea, Taiwan, and the United States.”
Most companies need to start at some point, as KPMG has observed. It is good to be ready for the transformation process. After all, CIMs could be the new future for not just machines but the humans and companies around them.
There is no point in spinning the wheel when the world is moving on a new axis. Better have these three wheels, than be a third wheel.
Author- Pratima Harigunani