Using software development to focus on opportunities, not problems

Register now

Industrial companies have long asked themselves big questions when it comes to innovation. How do we build a better jet engine? How do we create a safer, more efficient sub-sea system?

The core tenet in this question — how do we build — has become a cautionary tale in the age of IoT. It takes years to build (or rebuild) something from scratch. Before the cloud, data analytics and AI, this wasn’t much of an issue. It was expected. Armies of smart engineers required time to develop a product, and faced no major competition during the reveal of said product years later.

Today, technology evolves so rapidly and there are many options to iterate, that taking years to build something completely from scratch means it will be outdated by the time it's finished.

Now is the time to frame problems within big innovations as opportunities. An opportunity-driven approach means educating and iterating to add business value right off the bat. In other words: instead of getting bogged with the larger idea of how to build a better jet engine, ask yourself which services and data do we need right now in order to start building a better jet engine?

Old World Approach Versus New

The traditional, old world approach to development is rooted in a waterfall system that captures value in one to three years. Based on an independent data infrastructure for storage and a common taxonomy, this process focuses on milestones in lieu of product deliverables.

Connected equipment and systems are slowly integrated with experience and analytics to capture value. Waterfall presents a way of working that is difficult to course correct. It keeps your team working towards a stagnant goal while shunning flexibility. By allowing for learning along the way that can deliver deep insights into creating more customer-friendly and value-centric designs.

The opportunity-driven approach relies on agile, two-week sprints to capture value in cycles of 90 days or less. Made famous by software development, this methodology identifies a business process improvement, then develops the final product in sprints by defining the end product and accessing and transforming the specific data required for deploying the new application and capturing value.

These sprint cycles enable fast feedback, and quick wins that can be extended across a fleet or business area quickly. It is a way to scale and improve during the process of product development, rather than working towards milestones that may be less relevant in a fluid, dynamic business environment. It is also a way of ensuring that new technology is continuously integrated into products, rather than creating such products in a vacuum.

The Crucial Difference Between Building Turbines and Building Software

As for why building something takes so long, it’s useful to remember that great work often requires a large effort, sometimes in secret. While this still applies to heavy equipment like turbines, it is no longer true of software.

For example Facebook, Google and Netflix have built great software in many releases. In the past, Google was able to drive adoption of advanced analytics and machine learning across its business — not just into its products but even into the back office operations and other daily activities. At companies like this, machine learning models are becoming like the spreadsheets employed 20 years ago — easy to access and use, by increasingly less-specialized people.

Industrial businesses largely haven’t adopted an agile approach because everything needs to be safe and thoroughly tested. This is true of equipment and machinery, but software doesn’t quite have these same setbacks — and can therefore begin adding business value right away. One such instance of this could be in using a cloud-based predictive maintenance (PdM) solution.

In the past these sorts of long, complex infrastructure projects typically do not deliver immediate value, which is another way software differs. And it’s also only now being addressed by business because asset-heavy industries had such strong balance sheets that the urgency wasn’t there until the market conditions changed, such as oil prices dropping.

Getting Stuck

The tradition of waterfall-style product development means most industrial firms focus on addressing large problems. While this line of thinking can be productive to an extent, maintaining a focus on the big issues is easily derailed when an immovable hurdle is plopped in the way of resolution. If an insight pops up during the design process, the train is already moving, and adding innovation may risk delay or derailment.

Keeping pace with innovation means staying limber and spending time more productively. With waterfall, your company could be missing opportunities to optimize what’s needed to grow the business and improve efficiency. And in the era of IoT, if you take years to build a product, you’re moving too slow to take advantage of what you’re building. Technology may have already surpassed what you spent so long to build in the first place.

Cultural Resistance Despite Value

Value aside, industrial-type businesses still face challenges in the form of old cultural beliefs that threaten to hold enterprises back. For instance, it takes an IT department of 5,000 people to build a complex infrastructural backbone. If some of this backbone is provided as-a-service, the belief that this team would somehow become redundant is incorrect.

Rather than being automated away, those workers would have shifted priorities from building base infrastructure and server rooms to monitoring PdM solutions and addressing concerns before they become full-blown problems. Additionally, more security, architects and ops would be needed.

IoT is already making a recognizable impact on industrial businesses, which results in greater efficiency, lower operational costs, and higher revenues. Take for instance in agriculture, where driverless tractors can more effectively fertilize crops all while reducing the number of trips it makes. Or how sensors help to optimize irrigation and monitor soil composition to yield the most from a field of crops.

For reprint and licensing requests for this article, click here.