For the past few years, end of year predictions about supply chain innovation has focused on Robotics and AI (like this 2017 prediction piece). You can’t go to a tradeshow without buzzwords like “artificial intelligence”, “blockchain,” and “machine learning” being thrown out left and right.
But I’m willing to wager that the vast majority of the speakers I’ve heard at these conferences talking about the potential of AI, robotics, and blockchain still do most – if not all – of their day-to-day work inside a spreadsheet. Gartner’s 2019 predictions report put it very well: “Through 2020, 80% of AI projects will remain alchemy, run by wizards whose talents will not scale in the organization.”
Even if AI were mature enough for large-scale deployment in the supply chain industry, there are thousands of smaller businesses that don’t have the capital to invest in these kinds of solutions.
But affordability isn’t the only barrier to access. You can’t make good use of advanced solutions until you solve more fundamental problems with the way data flows in and out of your organization.
Most supply chain professionals need to focus on block and tackle fundamentals, not in lieu of considering AI solutions or instrumenting blockchain protocols, but in order to ensure that they are prepared to actually deploy the supply chain technologies and services of tomorrow.
Holder Says... |
 |
| Solving for block and tackle fundamentals is all about creating efficiencies and reducing supply chain days. |
 |
What do you say? |
 |
| Click here to send us your comments |
|
|
|
The Block and Tackle Fundamentals
The hype surrounding AI and distributed ledger technology (DLT) has had the unfortunate effect of de-prioritizing more pressing problems that can be addressed with straightforward, affordable software solutions.
Supply chain departments are taking on more challenging IT projects like omni-channel initiatives. At the same time, they are increasingly responsible for contributing to the overall profitability of their organizations. The conditions have changed, and the faulty IT infrastructure underlying many of today’s supply chain systems will inevitably buckle under the pressure.
If you haven’t done so already, addressing problems created by low-fidelity data and data silos between departments should be your first order of business. Solid supply chain operations are built on high-quality, high-fidelity data that flows freely between departments.
The good news is that the block and tackle basics of getting your data in order can largely done be addressed with affordable software that your existing workforce can immediately begin using.
Look for solutions that focus on eliminating data silos and that provide a unified view of your supply chain. Do your best to work with providers who can quickly integrate their solution into your existing operations and workflow.
Normalizing data and bridging silos between departments aren’t sexy problems to solve, but if they remain problems, thousands of companies will be left behind.
Successful business leaders – especially in the most technology-driven industries – don’t invest in innovation for the sake of being innovative. They innovate on outcomes. They leverage whatever tools and technologies will most effectively and efficiently close the gap between where they are and where they need to be.
Succeeding Today is the Best way to Prepare for Tomorrow
While the future of blockchain-based solutions and DLT in general remains uncertain, it’s hard to argue against the idea that every industry will at some point grow dependent on AI and machine learning in particular.
Addressing your organization’s immediate data fidelity and access challenges, is the best way to both solve today’s business challenges and de-risk future AI integrations.
Astonishingly, data scientists spend just three percent of their time actually building models; About 80% of the typical data scientist’s time is consumed by the tedious and decidedly less exciting work of preparing data. It’s no surprise that most failed AI implementations fail at the data-integration stage.
The fact is that AI models are only as good as the data we use in building them. It takes many (expensive) hours to effectively train AI models on learning data. While this is a gross oversimplification, the basic fact is that flaws in the learning data can compromise the integrity of a model. And you can’t just write a bit of code to patch that kind of problem. You pretty much have to start over. That gets expensive, fast.
So what can you do now? Start as you always should – with a clear business plan and clear business goals. Solving for block and tackle fundamentals is all about creating efficiencies and reducing supply chain days. What specific outcomes you are hoping to achieve by taking the next step? Take the time to better educate yourself about the kinds of AI solutions out that are most likely to meet your particular business needs: from smarter inventory management and predictive maintenance of warehouse equipment to insight into consumer expectations and trends and reducing costs per unit.
AI is not just an IT projection
While you should understand the basics of machine learning and blockchain protocols, what’s most important is that you adequately prepare for the operational and cultural changes that will accompany the adoption of these technologies. Adopting and deploying AI and other transformative technologies is never simply an IT project.
To be successful, you need people from across your organization communicating clearly and working toward the same goal. All too often – in their rush to innovate at all cost – companies make the mistake of overlooking this difficult work. Usually, things don’t go well.
Go get your data house in order. It’s the single best thing you can do right now to enhance your business outcomes. Then, when the time is right, you’ll be ready to adopt the AI solution that makes the most sense for your business, knowing that you’ve maximized your potential to get the most out of your investment.
Any reaction to this Expert Insight column? Send below.
Your Comments/Feedback
|
