Even as the disruption caused by the pandemic in the manufacturing and supply chain industries makes companies hold back investment in high technology; leadership teams continue to look for Artificial Intelligence (AI) to address their revenue generation and cost effectiveness concerns.
“A large number of manufacturing companies, and supply chains as a collateral, have had to halt their operations. Many SMEs and large manufacturing plants have halted or postponed any new technology upgrade in their factories in order to recover from the losses caused by the lockdown and economic slowdown,” says Sangram Kadam, Vice President and Head (APAC and META) at Birlasoft.
At the same time, 94% of the 200 chief executives surveyed in India by PwC in August-September 2020, affirmed of either having adapted or planning to adopt AI in their organizations. Talking to industry experts about challenges in adaptation of new technologies in the background of the disruptions caused by pandemic, some issues find common resonance.
Inability or inattention to quantify benefits of adaptation of new technologies in terms of ROI, unavailability of applicable data, shortage of talent, disparate incompatible nature of the data available, inadequate storage facilities and slow response time are some of the issues that find common resonance among industry watcher. Extent and pace of digitization in the country, inadequate availability and training of talents, privacy and security concerns and employee resistance to change are some of the other challenges, it is felt.
QUANTIFICATION OF BENEFITS
The early adopters of AI say the advantages of using this technology are immense as it allows you to forecast or predict when or if functional equipment will fail, hence, ultimately reducing downtimes. “Our manufacturing team is able to continuously operate and meet the production targets with zero breakdowns and minimal downtimes,” says Manish Jha, CIO, Addverb Technologies.
“To quantify, it could help us in increasing the performance metrics nearly 30% in terms of delivery schedules and the inventory management,” affirms Jha.
The belief is that any investment decision will essentially have to answer what expenses a business decision will incur and what returns it will generate. At any time, an entrepreneur has to weigh between different available options of investment contending for attention.
Votaries of AI, therefore, have to be able to project similar stories of quantifiable improvement in KPIs, pre and post adaptation of the technologies in the organization, keeping all other variables constant, feel industry experts. However, according to some others, there are tangible and intangible benefits to adoptions of technology and the intangibles have to be taken into account as well.
“The obsessive need for ROI justification for adopting any automation technology does not justify the case for automation to ignore parameters such as increased levels of personal and material safety, operational accuracy, real-time visibility for better planning and increased levels of job satisfaction among others,” Jha continues.
UNAVAILABILITY OF CLEAN DATA
The key facilitator to the success of Artificial Intelligence is access to data. An AI system has to learn from data for it to deliver on its function. AI builds and stores knowledge using data from multiple sources. “Access to clean, meaningful, high-quality data is critical for the success of AI initiatives, but can be a challenge in manufacturing,” notes Kadam.
According to him, manufacturing data is often corrupted due to numerous factors including the harsh operating conditions of a factory floor. Besides, historically plants have been built using proprietary systems that do not converse with each other.
Jha, in a similar refrain says, “AI needs large sets of data fed in an organized way. In case of mature processes and industries, there is no dearth of available data. But for the evolving processes and industries, it is a quite challenging task and leads to poorer accuracy of the insights and recommendations.”
Non-availability of data inhibits full realisation of its potential even where the technology is already being harnessed. “As we are in the process of establishing many processes and methods, we are still constrained by the availability of the data,” says Jha, who feels greatly benefited by the use of AI in production planning, inventory management and predictive maintenance.
“Although we are using the different features of AI in our software, we are yet to realize its potential in a full-fledged manner,” he says.
DEARTH OF TALENT
Talent for AI is in short supply, more so in manufacturing.
“The perception that the manufacturing industry is risk-averse, not sexy or cool makes it even harder to attract AI talent,” maintains Kadam.
But shortage of AI talent is not restricted to manufacturing.
Elaborating, Kadam points out that AI projects require multi-disciplinary teams with expertise in data management, algorithms, and machine learning.
“Most OT experts are not trained in building predictive models or using advanced analytics powered by AI. Instead, they rely on teams of data scientists and advanced analytics to assist them in using the latest technologies in operations.” “How can manufacturers hire AI talent when it is in such short supply?,” asks Kadam.
Communication & data modelling protocol Kadam points out a large number of machine tools and production systems that are used by the manufacturing sites. The technologies they use are often different and sometimes competing. Some of the software they use may be outdated or may not be compatible with the rest of the system. It is, therefore, left to the plant engineers to determine the best way to connect their machines and systems.
“The issue with interoperability has to be addressed,” says Kadam emphatically in this context. He favours an ecosystem that offers compatible components that use standard rules and frameworks to connect to ERP, MES, and PLC/SCADA systems to address the interoperability issue. “OPA (Open Policy Agent) UA (User Agent) is becoming the essential protocol for Industry 4.0 communication and data modelling” he says.
DIGITIZATION/CLOUD
AI is a data guzzling technology. The solution to meet the need is cloud computing. However, people in the know in the industry tend to veer around to admitting that India currently just does not have the storage facilities to keep pace with the exponential growth in demand for new technologies like AI.
NITI Ayog has suggested setting up of AIRAWAT, (AI Research, Analytics and knoWledge Assimilation plaTform) that will provide a platform for building infrastructure.
“Data lies at the core of development technologies.” These words of NITI Ayog CEO, Amitabh Kant, in the closing session of RAISE 2020 Summit gave an assurance in the industry circles of the cognizance of the matter in the highest circles.
INFRASTRUCTURE FOR REAL TIME RESPONSE
Cloud however, according to Kadam, has its limit. Decisions have to be taken in real-time, acted upon in a few milliseconds. Applications like predictive maintenance or predictive quality call for ultra-fast response.
“The system cannot wait for the round-trip journey to the cloud to perform data processing and get actionable insights,” Kadam elaborates.
It would be more efficient to process data locally near the source of data for faster response. That brings in edge-computing.
“It becomes more efficient to process data locally near the source of data for faster response. Realtime decision making and local control systems need edge-based computing,” maintains Kadam. “The ability to deploy predictive models on the edge devices such as machines, local gateway, or server is absolutely critical to enable smart manufacturing applications,” says Kadam looking beyond the cloud.
AI TO GROW
Lack of confidence in abstract algorithms behind AI technology among manufacturers, concern regarding data security, employee resistance to new technologies and the spread of digitization are some of the other factors flagged by industry seniors for the inhibition of adaptation of new technology.
Riddled though it is with challenges and the pandemic disruption, AI is the new technology the manufacturing industry is aspiring to. “Manufacturers now are leading the way in applying Artificial Intelligence technology, applying AI-powered analytics to data to improve efficiency, product quality, and employees’ safety,” concludes Kadam.