Artificial intelligence (AI), like 5G, often gets lauded as the answer to everything. Undoubtedly there are myriad potential applications for AI in the automotive sector. Whether it is a technology that resides in a vehicle controlling it autonomously or within business operations, but to date, the talk has very much outpaced the reality of deployment and the impact of AI.
There is a crucial point to understand when it comes to AI – it is not a ‘product’ you can’t buy it off the shelf and magically get the insights from your data your business needs. Or at least you can’t do that if you want to achieve the step-change in business operations that AI evangelists promise.
The 2020s are going to be a critical decade for AI. However, it requires the whole business, including finance, IT, product people, production line directors, engineers and development partners, to come together to deliver a solution. Without this collaboration, AI projects will only ever scratch the surface of any organisation and will not become the de facto approach.
The AI challenge
Let’s look at an example of what I am talking about here. Production, either in terms of managing supply chains or in quality control say, is one of the key areas where AI can potentially deliver benefits – informing decisions and realising efficiencies.
It sounds simple right?
For AI to make even a fundamental decision, you need to harvest a broad range of data – not just from machines and processes on the factory floor, but potentially from every part of the business, finance to customer data. And the fact is that historically organisations don't work this way.
The problem is that there are still too many legacy systems in place that keep data in siloes. Merely putting an AI layer on top of this infrastructure will only deliver cosmetic improvements, not the desired step-change.
Walking before you can run
Before even thinking about AI, automotive manufacturers need to go on a journey to become data-rich and insight-led. That journey will enable them to exploit the full power of AI.
The industry needs to break down those data siloes and use modern technologies and ways of working to surface the value in all that data. For example, without the right information on customers and customer behaviour how can automotive companies possibly make the right decisions about what car to build next, regardless of whether that decision-making is through an AI or human processes?
It is only once these building blocks are firmly in place that automotive manufacturers will be able to gain the insights they need to inform decisions up and down the supply chain. Think of the automation, rapid decisioning & insight you could garner from a mature data driven business.
And only once brands reach this point should they consider bringing in algorithmic automation to increase efficiency and explore deeper insights to decide what is the right vehicle to build, and ownership and usage model the consumer wants.
AI isn’t a short-term cure-all, it is just one part of a much more fundamental and involved transformation journey.