Enterprise software has been redone many times over with new technology, while the core design of business process automation has never truly advanced.Content Continues Below
Of course, enterprise software is more efficient today. We have added analytics and democratized user access. Many robust categories have developed, such as customer relationship management and talent management, to name just a few. Most of us probably can’t imagine our work lives without these applications and systems.
That being said, users manually enter data into enterprise software systems, triggering business process automation and execution. This is what SAP, Salesforce, Oracle, Workday and every other enterprise software companies have been designed to do — humans entering data.
From automating business process to automating decisions
Today, we’re at a seminal moment in the history of enterprise software. Machines, drones and robots are generating data that is not only significantly larger in volume, but arguably more impactful than that of previous generations. Whenever a Boeing 787 is in the air, for example, it generates a terabyte of data; the engine alone tracks data on 5,000 parameters. Drone fleets and intelligent, connected factories are also creating terabytes of data every day.
This volume of data presents incredible opportunity, and with the rise of sensor technology and IoT, the cloud, edge computing and, most importantly, artificial intelligence and machine learning, we can and need to move from automating business processes to automating decisions. A new kind of enterprise software is emerging as the industrial AI ecosystem takes shape.
Augmenting human intelligence
Artificial intelligence has a lot of hype right now, and in my opinion, it is for good reason. In every business, human intelligence and skill will be augmented and revolutionized with AI. Not in 10 years; it is happening now. And one of the greatest opportunities for software to reimagine business lies in the global industries that have become the infrastructure for the world and engine for economic growth.
What does this look like? An AI-driven business uses machine learning models to power and automate critical decisions at every step of the value chain, reducing cost, increasing safety, ensuring regulatory compliance and unlocking productivity at levels we could not previously imagine.
The impact of AI at scale is transformational and crosses industries. Rio Tinto, one of the world’s largest mining and metals companies, established a “mine of the future” program in 2008 to equip frontline employees with intelligent software to enable them to improve performance based on data and contextual knowledge, ultimately allowing for better, faster decision-making. Rio Tinto has one of the largest fleets of autonomous trucks, ensuring operational safety, reducing fuel use and increasing speed of delivery. Between 2015 and 2018, Rio Tinto predicted it would save up to $200 million annually because of its AI-driven digital transformation initiatives.
Industrial AI: A new breed of enterprise software
Industrial needs are changing, as is the world. In my conversations over the last decade with board members and C-suite executives from the world’s largest industrial companies, I consistently share three pieces of advice on using AI across a business to enable digital transformation:
1. Lead with outcomes.
If you buy Salesforce, it’s generally a great decision for capturing customer data, your sales rhythm and operations, but there is no guarantee that your sales team is going to perform well. If you buy SAP, it is good IT, but there is no direct line to your business running well. That isn’t good enough for enterprise software anymore, and particularly not for industrial businesses. You must demand outcomes that are real and can be measured — annual production in oil and gas and energy, velocity in rail; these KPIs matter to the business because they translate into dollars.
2. You need a new technology architecture to transform your business.
The future of industry requires an entirely new architecture that is cloud-native, but not native to any particular cloud. It must be intelligent at the edge, meaning devices, DCS, control systems and gateways. The architecture must be born in the era of machine learning, because the amount of data that machines generate is much higher at a much greater velocity and variety.
3. Content is everything.
Traditional enterprise software vendors have done lip service to verticalization, and OEMs in industrials have done lip service to operational domain expertise and openness. OEMs have a specific view, having manufactured the asset. But end users looking to do powerful industrial AI work need independent platforms with deep industrial asset data and content — historical equipment and process data, failure data, reportable conditions. This leads to insights across all machines and complete environments.
In the not-so-distant future, every industry will have its autonomous car moment. It’s happening now — the world’s oldest industry, agriculture, is developing model-driven automated farming to improve crop yields and reduce labor costs based on historical data, weather, GPS/telematics data, and field data. A single strawberry robot harvester has the potential to mechanically pick a 25-acre field in just three days and replace a crew of about 30 farm workers — and with AI, it’s getting smarter, faster and better every time it is in the field.
Sophisticated business leaders are recognizing the new industrial AI ecosystem. It isn’t a traditional enterprise software investment, because it isn’t just about process improvements anymore. There is a direct correlation between the software and the company making or saving more money. The real digital transformation happens when every one of these industries transforms its core offering with software. If you’re not doing that work, you’re not really doing digital transformation of industry.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.