Digital transformation takes up to three years to show results
Published on : Sunday 01-03-2020
What is the kind of impact emerging technologies are having on industrial automation?
Industrial automation is an on-going phenomenon from the past four decades. We have already seen programmed bots handling several tasks on the production line and a few systems replacing dangerous and dirty jobs on the shop floor.
Since the past five years, we are witnessing the arrival of intelligent systems – which can take some sort of decision based on the right kind of data. Technologies such as machine vision (MV) have automated quality checking, Industrial Internet of Things (IIoT) – a collaboration of sensors and devices over an industrial network has digitalised many operations and provided a clear path towards predictive maintenance. Drones have helped detecting leakage of dangerous gases, monitoring equipment. The cloud is helping storage, connection and distribution of data and providing several useful applications under the SaaS model. Robust industrial PCs, industrial networks are providing strong local infrastructure backbone to support modernisation of technologies.
Overall, I would say that industries have started feeling the presence and impact of emerging technologies. But, these are still early days of adoption in my opinion. The best is yet to come.
How is Artificial Intelligence helping the cause of fully autonomous manufacturing?
Both artificial intelligence (AI) and autonomous manufacturing are very broad terms. There are many intricate layers in both of them.
The four components of AI – the perception (enabled by machine vision and sensor networks), reasoning and planning (both enabled by the core algorithms and machine learning) and motion (enabled by robots) – could help realising the dream of fully autonomous manufacturing over time. But the AI would help full autonomy of manufacturing processes only if we have adequate quantity of high quality data about those processes.
The fully autonomous manufacturing process would contain all of the traditional manufacturing processes and a few additional processes related to data processing, machine learning and interaction with cyber physical systems (CPS).
There are few cases of partial automation but we have still not arrived at fully autonomous manufacturing as we are yet to establish standardised protocols and APIs for interoperability of different types of bots and traditional systems.
Despite the many advantages, there are trust issues when it comes to Cloud Computing. Are the fears exaggerated?
I have been advising MSMEs in technology adoption since 2009 in the UK and India. I see that the cloud has not come out of the dark cloud of mistrust despite showing a significant improvement over the years!
The concerns and fears of cloud consumers are somewhat justified because the cloud is a vast commercial collection of compute, storage and software, owned and operated by hundreds and thousands of small, medium and large providers under varying terms and conditions. No single vendor owns the complete cloud footprint of a cloud consumer.
However, the cloud is quite secure, technically. There are several types of security deployed on the cloud at several layers of the cloud. Your communication with the cloud is secure in most of the cases, thanks to SSL encryption. The virtual machines that host the cloud application and cloud databases would have their own access controls and often reside in secure datacentres. The cloud application could have role based security. A cloud application is quite secure if coded and tested as per OWASP guidelines. The interaction between the cloud application and the cloud database would pass through specific ports, firewalls and could be encrypted by the application itself. Most of the cloud databases have built-in authentication. Tables inside the cloud database can have row level authentication. Fields inside a row can be masked and encrypted. Nearly everything is logged. If you have budget, you could purchase line encryption from your Internet Service Provider (ISP) to ensure every keystroke is encrypted. Thus, there is a lot of security and traceability on the cloud.
However, the insecurity could be off-the-cloud, hidden in the small contractual terms and conditions of cloud usage. For example your SaaS provider may not agree to ensure secure back-up of your cloud data if you choose a provider and an agreement with some other provider!
Can a technology like Blockchain reduce the inefficiencies in manufacturing operations, especially in supply chain?
I think many of the analysts are getting carried away with the blockchain technology use- cases. Put in simple words, blockchain is a distributed ledger which records a transaction in a manner which could never be altered or erased once created.
To what extent the blockchain could help you depends on the intensity with which you would like to govern your supply chain. For example: If you are a manufacturer of diamond ornaments or diamond tipped drilling equipment, then you might need to watch your supply chain intensely because a slip, mistake, misinterpretation, miscalculation, wrong entries, delays related to diamonds could adversely impact your operation, finances and reputation! If you are in such a business, yes, blockchain would help you.
To store something as a blockchain, you need to encrypt the content, create blocks, spread them around the world and record somewhere the details of each block. To retrieve blockchain content, you need to locate each block, retrieve them and decrypt them before consumption. These steps consume energy as they are computationally expensive. Though blockchain isn’t an answer to all types of manufacturing needs and woes, many solution providers are force-fitting it into their solutions. The financial and environmental costs of using the blockchain technology should be justified by the use-case.
Is the pace of technology too overwhelming for most enterprises, especially MSMEs?
In my opinion, the pace of technology is overwhelming for everyone – including MSMEs. On top of the inherent complexities of a technology, we are adding layers of our own misinterpretations and generalised thoughts and opinions.
If you search online for information about any emerging technology, you would see many blog posts, promotional contents, articles, surveys, and academic papers, most of which are high level euphoric summaries.
To better understand a technology, we need real life case studies written – after the deployment of technologies and by the practitioners of technologies. Such case studies benchmark the real contribution of the technology to a business problem. Unfortunately such case studies are not written in all cases for many reasons including confidentiality.
The knowledge of technology is within agnostic consultants who could be expensive to hire. Thus, I am not surprised if MSMEs feel overwhelmed by the rapid pace of technological changes.
What should be the roadmap companies should follow in adopting these technologies?
I recommend three steps to companies that are writing roadmaps to adopt technologies:
First of all, understand and document your own business processes and your supply chain. If you understand the relative importance, complexity and risk of each of your processes, you would be able to judge whether, when and where you need to use an emerging technology. For example: Blockchain could be more useful to a diamond ornament maker as opposed to a washing machine producer.
While drawing roadmaps, make up your own priorities – for example – you could plan to automate those processes that are causing repetitive strain injuries (RSI) to your staff. Or you could start with the cloud – moving less risky data and operations to the cloud first. While drawing plans, you should explicitly mark processes that you are not comfortable playing with. For example: operations that involve your trade secrets and intellectual property.
Second – plan to fail fast. Select a few pilot processes and plan to welcome new technologies to automate them. Best way to do so is to run pilot projects. Such projects not only give you exposure and experience to the new technologies but also provide you an objective basis for decision making.
Third – plan to allow time for the technology to transform your enterprise. A rule of thumb is that a digital transformation takes up to three years to show results. So, plan to be patient. While you wait, plan to monitor the automated operations and adjust the technology or the business process or both as needed.
A typical roadmap would have multiple waves of piloting and adoption. And, there is no standard roadmap or a perfect one.
Mahesha BR Pandit, is an alumnus of MIT Sloan School of Management, Post-Graduate in Computer Science from BITS Pilani and currently a PhD Research scholar in the field of AI at Chitkara University. As a hands on technologist Mahesha enjoys building software solutions and publishing his thoughts. Since 1996, he has served global customers through his employers and guided digital transformation initiatives as a consultant to the UK Government’s Technology Strategy Board. Now he is the co-founder and CTO of Rhytify Technologies (https://www.rhytify.com) – a tech-start-up that provides AI software, review- as-a-service, cloud, eGovernance, and digital transformation and industrial automation products, consulting and professional services. The company has a research wing focusing on air-computing, bioinformatics, image-processing, model driven software engineering and robot ethics.Mahesha is reachable on [email protected] and [email protected]