As complex algorithms, machine-learning and artificial intelligence become mainstream, it seems inevitable that they’ll disrupt global supply chains. Companies, whether they’re Maersk or GSK, Jaguar Land Rover or Tesco, increasingly need distribution and inventory management systems that are self-learning, predictive, adaptive and intelligent; so-called cognitive supply chains.
“We’re at a unique moment in the evolution of the supply chain where advanced technologies have matured enough to match the proliferation of data. There is a current transition among leading companies where they’re aggressively reinventing from the inside out,” explains Jesus Mantas, managing partner for global strategy at IBM Global Business Services.
“The typical supply chain in 2018 accessed 50 times more data than just five years earlier. There’s an increasing focus on supply chains to reduce costs across incredibly complex, global operations. However, less than a quarter of this data is being analysed in real time.”
Cognitive supply chains are able to sense in real time, understand implications and trade-offs; they’ll also drive competitive differentiation to the next level
Intelligent inventory management systems still few and far between
The fact is smart, fully digitalised supply lines and inventory management systems that think for themselves are still thin on the ground. Amazon’s anticipatory shipping technology, which calculates demand for items in specific locations and moves them around efficiently, is still a widely used example. Yet it’s not the norm.
“There is a huge appetite to develop cognitive supply chains, which is only set to increase as more success stories come to light. But for many, achieving a fully cognitive one is currently just a pipe dream because of poor quality data,” says Alex Saric, smart procurement expert at Ivalua.
The age-old issue of siloed, unrationalised-data legacy systems, as well as companies that don’t talk to each other, is widespread. And with deploying cognitive supply chains, there is still a risk of being a first adopter, at the same time as there’s the worry of being left behind, while the dread of potential failure can be an even bigger limiting factor.
“We still see many companies stuck in pilot purgatory. They have many artificial intelligence learning pilots, as well as proof of concepts, but with few examples at scale,” says Matthew Burton, partner at EY Advisory.
Myriad benefits to cognitive supply chains
Get it right and the benefits are legion. Algorithms can tune supply lines based on human behaviour. The outcome can be better, personalised, customer service with lower inventories and a better utilisation of factory hours. Just-in-time supply chains prevalent in automotive, food and healthcare, which look to cut costs by reducing the materials a company holds in stock, are eyeing up these innovative solutions.
“A cognitive supply chain can help companies mitigate risks, improve insights and performance, as well increase transparency. Having one is crucial as the global trade wars wage on. Forward-thinking organisations that adopt these digital capabilities will be best prepared to navigate an unpredictable future,” explains Mike Landry, supply chain business leader at Genpact.
Growing global competition and market pressures are forcing companies across all industries to do more with less. As data explodes and becomes available from multiple sources in real time, it‘s also impossible for traditional supply chains to keep up.
“Companies can no longer use a historical statistical approach for demand streams to make sense of what the future will look like. Cognitive supply chains are able to sense in real time, understand implications and trade-offs; they’ll also drive competitive differentiation to the next level,” says Fred Baumann, general vice president of industry strategies at JDA.
Investing in a cognitive supply chain doesn’t need to cost the Earth
Building cognition into supply chains and inventory management systems doesn’t have to be a multi-million-pound, insurmountable corporate mountain. Buy-in from the C-suite helps, so does a bite-sized approach.
“Companies should start with simple use cases as part of an overall roadmap. Starting with cost-savings targeting near-term return on investment to prove that its value can help spark change. The emergence of a new role, the supply chain architect, can also help drive change,” says Kevin Doran, managing director at Accenture Strategy.
The primary challenge once a business decides to deploy a digital supply chain is that data across the distribution lines, end to end, is not owned by any single entity. Data definitions are different across industries and at different stages.
“Consensus on data standards takes time and that process unfortunately doesn’t move as fast as the speed of supply chains. Companies need an open API [application programming interface] that can take digital streaming from any source and is capable of ingesting real-time data. No single company can achieve end-to-end visibility, as this also requires a partner ecosystem,” says Mr Baumann.
Cognitive supply chains are vital for the future
As technologies continues to mature so will the roles for management of the digital supply chain, as well as new concepts such as the virtual supply chain or digital twin. Dummy data can now be used to test issues with supply, distribution, inventories and products. It is based on a transparent supply chain strategy, comprised of rules on how to absorb and refine costs.
“Uncertainties such as pending tariffs or disruptions can be run through ‘what if?’ scenarios to understand the service, cost and risk implications of decisions and unexpected market conditions,” says Mr Landry.
Blockchain-enabled smart contracts also offer a particular advantage when it comes to cognitive supply chains. The reliance on extensive paperwork is still one of the biggest challenges. “By coupling smart contracts with other technologies, like the internet of things and tracking tags, smart contracts could automate individual processes within a supply chain,” says Ian McKenzie, associate director at Osborne Clarke.
In the United States, Walmart now requires suppliers of leafy salads to use blockchain technology. Romaine lettuce contaminated with E. coli prompted this. Using blockchain, Walmart tracks each item as it moves along the supply chain. This enables the retail giant to pinpoint the origin of contaminated food in minutes rather than days, with the potential to save lives. This issue has made the supply chain smarter in the process.
According to IBM, which is rolling out their Watson Supply Chain, we also shouldn’t forget the human element as we development cognitive systems. “Artificial intelligence allows organisations to free up their professionals from mundane, transactional activities to focus on higher-value responsibilities,” says Mr Mantas. This is perhaps where the real advantage lies.
Brexit rethink over just-in-time supply chains
Brexit remains a major unknown. This is especially true for complex, smart supply chains that span out across Europe supplying the UK daily with goods in many sectors from automotive parts to drug or food ingredients and aerospace components.
How supply chains are optimised comes into question when there’s a risk of fresh border checks or new routines, which could cause long delays at ports. Certainly, agile distribution lines could suffer.
“Self-organising cognitive supply chains often rely on a steady flow of just-in-time inventory. These will be particularly susceptible to disruption unless companies plan ahead and build additional inventory capacity to buffer for potential additional delays on imported or exported goods,” says EY Advisory’s Matthew Burton.
Tying up capital in inventory will be an issue and risk management will become vital if the new normal is volatility. The predictability of post-Brexit trade and how this can be modelled is a big challenge.
“Brexit should present the stimulus for companies to develop truly resilient and responsive supply chain management capabilities, underpinned by cognitive, data-led decision-making that senses and reacts to supply risk and disruption before it impacts the customer,” says Kevin Doran at Accenture Strategy. Not a big ask then.