An e-interview with Jagdish Upadhyay, Director, Global Product Marketing – Hitachi Insight Group.

Jagdish Upadhyay
  • Client

    Hitachi Insight Group.

  • Services

    Internet of Things (IoT)

  • Technologies


  • Dates



It is possible to equip old machines in good running condition with sensors to make them more productive. Is this advisable, how costly is this exercise, or is it better to opt for new machinery?

How about a Fitbit for machines – a typical Fitbit has 7 or more sensors – GPS, 3-axis accelerometers, digital compass, optical heart rate, altimeter, ambient light sensors, vibration motors in a tiny form factor that can run for days together with a single charge, and synchronises with mobile device app which pushes the information to a cloud based analytics software to provide your custom dashboard to see your health parameters. It’s the same idea behind the Industrial IoT add-on devices.


Legacy hardware assets are prevalent across all industries and any potential IoT solution must include provisioning solution for these assets. Some of these assets don’t have any sensor mounted – these are pure mechanical assets and not digitally connected with any infrastructure. Then there are other legacy assets which have sensors but communicate over legacy operational technology protocols.


In the case of mechanical assets, the question really is how much retrofitting is needed? It would depend on what we are trying to solve for. There are commercial off-the-shelf add-on products available today which can be externally mounted – just like my earlier example of Fitbits. These add-on devices come with pre-installed sensors and network connectivity to sense and capture wide range of sensor data such as temperature, pressure, humidity, vibration, etc. This data is then made available to on-prem or cloud based data science solution.


In case of legacy operational technology, there are several edge and fog based solutions that provide protocol translation and data management capabilities. Since there are custom requirement, the type of add-on device and size of integration determines the overall RoI from the investments.


What are the best ways of integrating old machines into modern concepts like IIoT?

Even though these assets may be old but may still have remaining useful or productive life. Therefore, we have to take an approach of retrofitting these assets based on scope of the solution. In case of hardware assets with legacy or proprietary network communication, there would be need for a gateway type device to fill the gap of protocol translations and data conversion.


There is strong business case for providing just the asset visualisation capability – just by capturing basing operating conditions, we can ascertain performance and utilization of these assets. While analytics helps in augmenting better decisions, its application depends on the use case.


How to simplify the IoT deployments?

The first step essentially is to establish the business case and identify the right stakeholders early in the journey. This may seem like a very elementary step but most projects get delayed or fail to deliver desired result due to lack of clarity in expectation from various stakeholders. IoT projects generally touch upon every aspect of business starting with IT, Production, Maintenance and Corporates. Therefore, it’s not just one entity which can drive this initiative.


The second step is to find a partner in this journey to help at every step of execution. Working with partner helps in gaining access to industry leading practices and not getting blindsided by internal processes and point of views. While selecting a partner or vendor, validate your vision with their capabilities. Many vendors are open to creating joint PoCs and pilots to show proof points before starting a full-blown implementation.


The third step is selecting right set of products and tools. Even though project decisions are biased towards the current needs, but the approach should be to factor in future vision of business and plan for scaling up or down as required.


The last step is to make sure there is a communication and change management system in place. Automation generally is mistaken as job redundancies and therefore it is important to communicate with employees and prepare them for the change. Automation helps in removing manual bias and therefore making processes perform better.


IoT is a journey and you would need a right partner to help you navigate. Please consider all aspects of your business needs including impact on ongoing production, existing operational technology setup, needs for data security and availability of key stakeholder throughout the project.


What would be your advice to SMEs in India who are reluctant adopters of automation?

There is a reason for reluctance across the industry – it’s not unique to India in my opinion. As per a recent survey by Cisco, it was reported that three-fourth of IoT projects are failing. It was not for lack of trying but there were other reasons such as collaboration between IT and Business which I talked about earlier. The other reasons were related to the culture of the enterprise and approach taken to deploy IoT solutions. The stakeholder alignment is key to successful implementation. The gaps in IoT expertise can be addressed by a partner when selected right.


So, it really narrows down to two key steps – identify the business needs and select the right partner before starting this journey.


(Jagdish Upadhyay is Director – Global Product Marketing with Hitachi Insight Group. He is a senior executive with proven track record in launching new Product offerings in M2M, IoT and IIoT, opening new market segments, establishing strategic vision, and developing successful new business opportunities.)