Choosing the Right Team to Succeed in Industrial Analytics Initiatives

Thinking about building your industrial analytics app in-house? To succeed you need the right team, and strong IT-OT collaboration.

In my previous post about Build vs Buy, which discussed when it makes sense to design your own analytics application or buy an existing product, I had mentioned that for the Buy strategy the right team needs to be in place. In this post, we will talk about the team structure and more closely examine the different types of data scientists emerging in the industry. The team needs to be a mixture of IT (Information Technology) and OT (Operational Technology) for the initiative to succeed. One cannot operate without the other. A clear corollary is that if IT is the heart then OT is the brain. The following roles are essential for the success of the analytics initiative:

1. Business Unit Head (OT): The budget holder and owner of the initiative from the OT side. He is the CEO of the initiative and should be the one who defines the objective and final business outcome for the success of the project. He should also conduct the final impact analysis so that he can become the evangelist of the initiative throughout the rest of the organization.
2. Chief Digital Officer (Chief Intelligence Officer) (IT): He is the topmost contact point from the IT perspective and is responsible for the final success of the project from a digital transformation perspective. This means that he will work closely with the Business Unit Head and ensure the success of the initiative from an executive standpoint.
3. Program Manager (IT-OT border): The program manager is essential for ensuring the coordination between the groups. He has to ensure that the engine is constantly rolling and both parties are equally engaged to solve a common problem rather than going down their own path. He is like the traffic cop who ensures that traffic is flowing the way it needs to.
4. Subject Matter Expert (OT): The SME could be from any team depending on the way the problem has been defined. If its an asset-related issue the SME would be a field engineer or plant manager. He is more hands on with the data and can guide the IT teams by providing the right input for analysis.
5. Application Data Scientists (IT): The DS team is responsible for interfacing with the product teams to ensure that everything is in line with the final outcome. They are hands-on with the product, and understand the algorithms selection, tuning and data manipulation needed to get to the end result. Application data scientist are masters at applying existing algorithms and strategies to solve the current business problem as compared to creating algorithms from scratch like the research data scientist.
6. Data Engineers (IT): The final heavy lifting of defining the ETL activities for the products to work properly would be done by the data engineers. As much as we want everything to work like magic, data preparation and pre-processing still remain the number one time consuming activity for any analytics product. Rather than running away from it, its better to address the elephant in the room upfront and have a specialist to tame it.

The reason I have defined the above hierarchy and roles for the Buy strategy is because there is no magic bullet in the market which will solve all problems at the click of a button. There needs to a clear intent from the side of the organizations which are planning to implement analytics solutions. That intent can only be seen if there is a right mix of IT and OT and both parties have some skin in the game for the final success of the initiative. One without the other is like taking away the brain or heart. The body cannot function without either.