Wide and Deep types of projects for industrials
- 3 minsIntroduction
I’ve delivered many projects for industrial enterprises. The two extreme case types are what I call Wide and Deep projects.
Wide projects are about building infrastructure and indirect effects, characterized by:
- A large number of signals, devices, and visualizations
- Different types of devices and connectivity
- A concrete scope
Deep projects are about the necessity to achieve specific direct effects and are characterized by:
- Extensive R&D and experiments
- Cutting-edge technologies
- Uncertain results
The two types of projects have different problems and methods and tools for delivery. Let’s try to uncover them.
Operational Excellence for Wide Projects
When you’re building infrastructure projects, typically, the final result is clear:
- N devices connected to the system
- M signals collected in the database
- X diagrams created and linked with data
- Y report templates built
- Data transferred to Z systems
Of course, many problems and strange requirements arise. But they all have solutions and can be resolved in various ways with different costs, for example:
- Imagine there are 20 sensors from a vendor installed, and two sensors have legacy firmware. Options include upgrading the firmware, replacing the sensors, or developing a connectivity module in your software for the legacy firmware.
- Or communication issues with customer staff occur. I bet you will find an approach to the right person. It’s about having the right people, communication frameworks, and project management experience.
Building large infrastructure projects with lower costs requires a lot of junior-level staff. They should complete a huge number of simple tasks like filling forms or Excel files with configurations for connectivity modules and drawing diagrams. The project delivery process should be supported by experienced professionals who prepare well-defined and concrete scopes for such tasks. This senior staff should accept results and maintain the right level of quality.
Summing up, you understand the scope, slice the project, and try to stay on schedule. The main goal is to finish the delivery before the deadline with the desired quality. Here, operational excellence is paramount, and approaches like agile practices and autonomous teams will likely obstruct and introduce unnecessary complications.
Agile and Experiments for Deep Projects
In contrast, when you build a solution that should bring direct effects (specific percentages of energy efficiency, total throughput, quality, etc.), the final result is not clear at all.
Even if you have delivered the same solutions earlier, the level of customization will be high for the new customer. Something that brings a 3% increase in energy efficiency somewhere could bring only 0.01% at the next site. You have to continue experiments and solve conundrums all the time.
You might encounter pitfalls like:
- Dirty and vibrating environments for computer vision
- Poor sensors for predictive models
- Lack of sensors for advanced control
- 90% of time spent with zero effect delivered
Sometimes, bypassing these issues requires many attempts and much effort.
To complete deep projects, only top-rated specialists are required. Even if the team has juniors, they should possess great skills. Problem-solvers are essential. Teams built with such workers are very autonomous and need just a problem at the input and some time to solve it. Here, agile techniques and non-standard project management ideas are necessary.
Conclusion
Anyway, both types could practice a hybrid approach with an appropriate ratio of methodologies.