- By R Narayan
Artificial intelligence (AI) has taken the world by a storm, owing to its fail-proof facilities that have made both life and business easier the world over. Realising the significance of Industry 4.0 has on business countries around the world are becoming increasingly aware of the potential economic and social benefits of developing and applying AI. India, not too be let behind is fully committed to providing impetus to AI and developing as a mode of production to be leveraged to alleviate its economy. The government has even set up four centres for promoting industry 4.0 across the country.
The applications for AI for a manufacturing SME can be for advertising, marketing, customer engagement, customer profiling, segmentation, predictive maintenance and several more. The applications do have a twang of novelty to them but their adoption and implementation remain a challenge owing to certain limitations in the framework.
Availability of Correct Data
"Data is the new gold" and unfortunately the Indian MSMEs still suffer from massive data poverty. Data availability and management can make or break an AI ecosystem. While Government platforms like GST and organic percolation of digital platforms have accelerated digital data on financial behaviour, services like Maps and GPS are deployed for easing logistics is also a rich source of data. Similarly, we can assume that some MSMEs that have modern equipment would be rich in machine logs. These are a step forward in data capture. A lot more needs to be captured for a meaningful AI problem statement to be solved. To fix this, SMEs are advised to start investing in data capture. It’s both culture and digital transformation journey.
SMEs need to adopt lower versions of enterprise resource planning (ERP) platforms and get their teams to adhere to a process that implicitly collects data for future leverage. Young companies have started making inroads into IoT and this will also enable detailed data capture, putting in place the foundation for AI in the coming decade. SMEs need to be enthusiastic about collecting data about their customers as many lack data richness for some facts and hence cannot use the predictive power of AI for timely resale or servicing.
Availability of AI Solutions
AI has become a bedrock for well-funded companies and remains a distant dream for SMEs who continue to expand their bandwidth on basic working capital management to run and expand the business.
The enabling AI expertise lies with the few who are already vendoring their services to organisations with deep pockets. Cost-effective cloud-based SaaS ML tools are available in abundance now but these have only mitigated the need for an AI programmer, not an AI expert.
These frameworks need to get more domain-specific and problem-specific. For example, rather than having a generic Decision Tree tool or a CNN tool, the tool needs to target a specific business issue such as- "Improve your FMCG product volume apportioning for a geography" or "Pricing your maintenance service for compressors and motors" or "Optimising the input material for your galvanized pipes based on target application".
MSMEs, for now, are using chatbots on their websites & the digital marketing tools offered by social media that are powered by continuous learning. This exposure helps diffuse the knowhow and the possibilities to the MSMEs for whom implementing anything more sophisticated is beyond their skill set as they're busy in improving their business, and developing improved distribution networks.
Scope of ROI from AI
AI problem solving is an expensive initiative. Getting ample data, putting it into a usable form and availing expertise on an AI application is a technically intricate, time-consuming and costly journey. In other words, as the economics stand getting a positive ROI is hard to get to in case one attempts an AI project at a small scale. This will get cash accretive as first, the price points of solution providers come down over a period of time; second, data capture via automation becomes a defacto practise; third, industry-focussed plug and play sub solutions are available for rental (SaaS); and fourth, anonymised data sets are aggregated for faster ‘data lakes’ and effective learning.
The Need of the Hour
As presented above, AI is not very inclusive of manufacturing MSMEs and only targets large industries in retail, FMCG, finance, telecom, healthcare, travel, and entertainment etc. The government must consider bridging this inclusion gap by undertaking awareness initiatives and mandating industry data and integration standards. It must introduce "data aggregators" whose main focus is to collect and validate data in multiple sectors and make it available to MSMEs at subsidised rates to facilitate data input for AI solutions.
Financially the government could provide tax breaks to SMEs investing in AI projects or to firms offering/building AI platforms for SMEs with a goal to keep pace with nations like China, who is steadily moving from investments in basic enabling infrastructure to AI infrastructure for the majority of the spheres of the country. Providing such impetus to AI for MSMEs will not only democratise AI across industrial sectors but will also accelerate productivity and efficiency in Indian MSMEs whose contribution to the GDP will finally be able to match the global average of 49 per cent.
(R Narayan is the Senior Vice President at FICCI-CMSME and Founder & CEO at Power2SME. Views expressed are the author's own.)