Unlocking value


Vijay Jaswal, Chief Technology Officer of APJ&MEA, IFS discusses the company’s commitment to helping organizations resolve their productivity, predictability and agility issues with their solutions

Elaborate on how the theme of ‘unlock business value, productivity, predictability, and agility’ at the recent IFS Connect Middle East underlines your customer focus.

The theme underlines the importance we attach to providing value to our customers. IFS is committed to helping organizations resolve their productivity, predictability, and agility issues and ‘unlock business value with Cloud and AI’. We provide that value by giving our customers tools that make their lives easier and their operations smoother.

We only sell to key industries. These are aviation, defense, manufacturing, telcos. energy, oil and gas, utilities, engineering, and construction companies, and any services-rich company. Companies in all of these verticals have one thing in common, they all have physical assets, be it an engine, a piece of machinery, or a pump in an oil and gas facility. Our technology ensures that all these assets continue to function smoothly, adding value and keeping our customers happy. If they’re not working, they add costs unnecessarily and customers are unhappy. All our technologies are around asset management, field service management, ERP helping ensure productivity and efficiency. We remove all wasteful activities from the processes and automate efficient processes that helps the user.


Elaborate on the new AI capabilities of your platform?

In our up-and-coming release, 24R1, we’re releasing a load of AI capabilities, not just at the front end and not just having copilots, but all the way through.  We have got various layers around data, data orchestration, bringing different data sources together. Data governance is a key focus area of AI, ensuring data is relevant, data is true to fact, not copyrighted or belongs to somebody else. The copilot capability is something we’re releasing later this month. If a user is using the system and wants to find out some information about the process that they are in, or the functionality of the system, historically, they have to find the correct document, the correct manual, or the correct PDF document or go on the web, and  that will consume a lot of time. Now all they must do is just write into the copilot, asking an explanation of the manufacturing processes, and the copilot will access multiple sources, user manuals, technical manuals, and even our online community portal and provide that information within seconds. The user productivity is thus enhanced. While that is from the user’s point of view, at the backend we have been using AI for several years now in scheduling and optimization capabilities. For instance, it could be the ability to ensure that when you have many engineers, they turn up at the right place at the right time with the right tools to fix the problem, whether it’s at a customer site, or a production line, or in an oil and gas facility. So, we have  had this AI capability for several years, and we have only enhanced it even more and added the extra capabilities around simulating possible outcomes.


What’s the feedback from your users around the scheduling optimization features?

The feedback has been amazing when it comes to the scheduling and optimization side of things because a lot of our competition doesn’t do this optimization in real-time. For example, if a telco engineer has for instance confirmed an appointment to be at your house, the next day between 9 am and noon, to install your new ADSL router and you have taken a half a day off, but no one turns up, which could be quite frustrating. So, if our customer were the telco provider, and you were the customer of the telco organization, in this case, if the engineer was perhaps delayed because of some traffic issue or was delayed at the previous customer site or any other issue,  this customer issue can be fixed with the IFS Scheduling Optimization engine. It can in real time deduce that an engineer is going to be late to visit you and then find another engineer that’s close by. It won’t be a three-hour time window for the scheduled visit, it will be more like 15 minutes or half an hour and the engineer will come between 10 and 10:15 am. And if you are able to track where the engineer is, you can even go out and get a quick cup of coffee since you know when the engineer is going to reach. This has helped enhance the productivity and efficiency of our customers and so feedback has been tremendous.


How does this take care of cost management concerns, and enhance sustainability?

To answer this, let me elaborate on my earlier example of the telco engineer. There are a lot of variables in the example. If an engineer turns up but doesn’t have the right skill set, or turns up with the wrong parts or the wrong ADSL router, it is a wasted journey. We’re big believers in increasing the first-time fix element of an engineer’s work. So, the engineer needs to go only once and that’s one side of sustainability. The other side is  one of our capabilities called remote assistance. In this case, before an engineer turns up at a job, the customer can use the phone and take a picture of the device that’s not working and the engineer can advise the customer on how to fix it. Or if an engineer does turn up but isn’t trained enough, rather than going back and bringing in a more experienced engineer, that engineer can just call the experienced engineer, and again, with a remote assistant, it’s like augmented reality, the experienced engineer can override and instruct what to do and that goes a long way to help from a sustainability perspective. And in our new release that’s coming out this month, we’ve also made allowances for electric vehicles, because optimizing and scheduling, electric vehicles versus petrol vehicles is different. While refuelling of a petrol or diesel car takes one to three minutes, recharging electric vehicles takes half an hour, if not longer, depending on the range. We make all these allowances into our system pre-builds. And then we can also do things like measuring the carbon footprint, measuring the emissions, and more, conforming to ESG objectives.


How do you work with different industries? What kind of level of customization goes into taking these solutions to them?

We have solutions for each industry that we focus on. And we work closely with our customers because who knows that industry better than them? For instance, we have a global customer event coming up in October later this year in Florida. And as we speak, we’re working with several of our flagship customers to create AI use cases for them. So be it maritime, telcos or oil and gas, we are working with our customers, and they’re liaising with our R&D teams to build these use cases, the caveat being they must speak at the event. And those solutions will be made into templates for those specific verticals. As we focus only on a handful of verticals, we can create to-the-point solutions that are relevant and will add value to that particular industry.


How important are the ethical considerations factored into AI models? What determines the precision of AI forecasting?

There are so many debates around the ethical side of AI. There are many facets to ethics and AI. The quality of data is one such aspect. The ownership of data is another – who owns the data, are we allowed to use that data or is it copyrighted etc. There have been articles about people complaining that some AI engine was stealing the thesis they did at a university and so on. There are several angles and also ethical concerns in terms of AI learning models. For instance, a previous company I worked with, has facial recognition software that determines your gender and estimates how happy you are and how old you are. We showcased it at GITEX but it wasn’t accurate because the model had trained on European faces and wasn’t ready for the multi-cultural environment of GITEX. This endorses the fact that the training of the model is also very important.  On the data side, there are several layers and one of them is the governance layer that goes a long way to ensuring that the data that goes in is relevant, is correct, isn’t copyrighted etc. Because once the data goes into the community website, you need to make sure that the data that’s coming in, is correct, that solution has to be correct because other users will see that. It is a long-running debate and by no means is it solved.


How are you breaking down data and functional silos in the core industries of your focus?

For an organization to be truly digital, islands of information have to be removed and backed by islands of information. For instance, if Application ‘A’ talks to Application ‘B’, then you may have Application ‘C’ as well. Sometimes users have to log into application A to pull out a reference code and use that reference code as a query to Application B, so that takes time again. What we try and do is ensure that all these different data sources are integrated so that the system does the hard work of pulling information, reference number from system A and use that information for system B, and all that the user sees is the response. That’s how it should be. But a key challenge that many customers have is ensuring all their data is accessible and accurate. IFS Cloud has over 5000 APIs for integration purposes. We also partner with a company called Boomi which is an enterprise integration customer. So, we can use this backbone to integrate all the various subsystems. It’s important that to be truly digital, you need to be truly connected. And we’re big believers in that.


What is further on the immediate roadmap of IFS?

There would be many more vertical AI use cases from a copilot perspective. As I mentioned, we’re releasing IFS copilot in a few weeks. We want to extend that capability to third-party co-pilots. So, for instance, in future releases, if a user needs to find out what the production schedule is in the factories today, they can do that from within teams or PowerPoint, rather than having to go into the IFS environment. We’re looking at enhancing our anomaly detection capabilities around asset management. We already have IoT or sensor-enabled assets, and predictive maintenance capabilities. But we want to push it out and improve it even more. You can expect a lot of enhancements on the AI front and also the vertical side of our industrial AI.


What is the readiness across industries and your customers to adopt such advanced AI solutions?

Some industries like telcos that are more innovative savvy are quicker adopters than some of the other industries. It varies from industry to industry and also differs based on the size of the organizations as well. I think the smaller family-run businesses are probably more cautious than the larger organizations that are a bit more innovative, and they want to have that kind of competitive edge by being early adopters. AI is  moving rapidly from the PoC stage, which is where we were last year, to realizing value and benefits.



Leave a reply