Unlocking the value of supply chain process through AI
Over the past four decades, much of manufacturing production worldwide has been organised in what has become known as Global Value Chains (GVCs). COVID-19 has struck at the core of global value chain hub regions, including China, Europe and the US.
Now, as production comes back up to speed, companies are scrambling for a means of monitoring the inbound flow of product, to figure out how it can be received, stored and shipped at a time when demand for all but the most critical items has ceased. This has prompted many businesses to review their current supply chain processes and to evaluate how they might build in resilience ahead of any future disruption.
Digitalisation is essential for industry in this current climate, both to increase margins and operational performance in good times and to adapt in the bad. AI has reached a key juncture where the real-world benefits are instantly recognisable.
Understanding where AI boosts existing processes
In the industrial sector, AI application is supported by the increasing adoption of devices and sensors connected through the Internet of Things (IoT). Production machines, vehicles, or devices carried by human workers generate enormous amounts of data. AI enables the use of such data for highly value-adding tasks such as predictive maintenance or performance optimisation at unprecedented levels of accuracy. Hence, the combination of IoT and AI has begun the next wave of performance improvements, especially in the industrial sector.
Furthering this automation, AI uses the historical IoT data to analyse trends which can help in streamlining and improving the supply chain process through cutting-edge solutions such as AI-driven operations scheduling. This provides recommendations to humans as to the optimal scheduling sequence, substantially reducing error and inefficiencies.Further, this AI learns as it goes and tailors its guidance to particular situations, gaining intelligence the more it runs. In addition, AI-driven robotic process automation removes the human element from repetitive tasks in varying levels of complexity, thus furthering increases in efficiency and accuracy. Through integrated workflows, much of the supply chain process can be intelligently automated. These types of AI-driven capabilities have the potential to redefine the business supply chain process. In certain industries such as oil and gas, these types of operational gains are more important than ever in today’s climate.
Early adopters of AI have deployed these technologies on-premises, in the cloud, at the edge, and through many types of hybrid architectures. AI itself is not one thing but comprised of several technology types, including neural networks, deep learning, natural language processing, computer vision, unsupervised machine learning, supervised machine learning, reinforcement learning, transfer learning, and others. These various types of AI are applied in different ways throughout the industrial world to create targeted solutions provided as descriptive, predictive, prescriptive, and prognostic analytics.
Overcoming the fear of automation
Beyond deciding where and how to best employ AI, an organisational culture open to the collaboration of humans and machines is crucial for getting the most out of AI. Trust is among the key mindsets and attitudes of successful human-machine collaboration.
Here are some practical steps to consider if a business is looking to explore the implementation of
Artificial Intelligence or Machine learning capability into their business process:
- Leverage AI to gain significantly more value out of existing industrial software: SCADA (an acronym for Supervisory Control and Data Acquisition generally refers to industrial control systems) and other types of control systems have become standard practice in most industrial facilities. Real-time and historical data is typically used for trending, reporting, and HMI visualisation. AI allows companies to get much more value and insight from this historian data through state-of-the-art technologies such as multi-variate machine learning and deep learning. By integrating software infused with AI into existing industrial IT infrastructures, businesses can greatly amplify the value and ROI by detecting and solving operational and maintenance issues before they become larger problems that often result in unplanned downtime. This alone can increase uptime by 10% annually, resulting in substantial avoided costs and efficiency gains.
- Allow AI to integrate into the core of the supply chain to take advantage of cutting edge capabilities: AI-driven operational scheduling and work process automation can eliminate mistakes and allow industrial companies to get the most out of the resources they have available. Supply chain success is critical to overall business success, and an increase in efficiency can often be the difference between turning a profit or not. AI provides incredible value in this area, and businesses shouldn’t be afraid to leverage its power as an integral part of their supply chain process.
- Use the cloud to ease the implementation of AI, allowing companies to scale: Artificial Intelligence is fast becoming the brains behind the cloud. Consequently, companies can quickly deploy and access a variety of industrial software capabilities that are driven by various types of AI technology. The cloud is the delivery mechanism, and SaaS is the commercial model; however, AI drives much of the value gained. Now more than ever before, AI is becoming more easily accessible and more cost-effective to deploy into industrial environments.
- Bridge the gap between AI and humans: In order to glean maximum value from AI, companies must ensure that they bridge the gap between AI and human understanding. A significant portion of the workforce today is somewhat distrustful or fearful of AI. It is critical that companies do everything they can to ensure that the benefits from AI-infused software are translated into the vernacular of the targeted worker. The benefits provided by AI must be put in context, be useful and actionable. If this does not happen, then much of the value of AI is wasted.
- Be open to continued innovation and change: AI capabilities continue to evolve and improve. Software will become more intelligent through combinations of AI capabilities in order to achieve more sophisticated machine-based thought and reasoning. Amid these changes, companies can reap more and more benefits through deeper insight into cost vs risk decisions, an improved understanding of business processes and associated efficiencies, and better forecasts of future events. By continuing to plan for and incorporate change, companies can take advantage of ever-improving AI capabilities and insight.
Time to reflect and evolve
Businesses now need to be incredibly agile to manage the costs of turning down production, followed by the working capital constraints to then rebuild production levels as economies recover. We are also seeing a period of distrust and disinformation while global supply chains are disrupted, data is key to traceability and provenance ensuring that drugs and food come from authentic sources. Better visibility allows us to understand where resources such as food and pharmaceuticals are and how we can get energy efficiently to those who need it.
Digital transformation stands to provide an immediate and compelling competitive advantage for those quick to adopt – and to demonstrate provenance. Businesses require intelligent software to address industrial pain points for value creation, productivity improvement, insight discovery, risk management, and cost optimisation. AI is a key differentiator and a propelling force behind improvements in the supply chain.
Biden’s Cybersecurity Executive Order; A Leading Solution?
Colonial Pipelines, Microsoft Exchange, SolarWinds. What do those names all have in common? They’re all subject to highly sophisticated breaches of the United States’ cyberinfrastructure. Each serves as a sobering reminder that, in an age of advanced technological power, criminality and danger lurk around every corner in an increasingly virtual, data-dependent world.
There has been a myriad of cybersecurity incidents in recent times, allowed by insufficient cybersecurity defences that have left both the public and private sectors vulnerable. As an outsider looking in, it’s evident that both sectors must prepare themselves for a long, drawn-out battle against increasingly complex and malicious cyber attacks from both rival nations ─ China and Russia, for example ─, and global cybercriminals.
The Cybersecurity Executive Order
To that end, American businesses and institutions likely breathed a sigh of relief at the back-end of last week.
Following on from President Biden showing his government’s willingness to, at least, attempt to mitigate future risks by approving America’s Supply Chains review, the ‘leader of the free world’ has signed yet another Executive Order ─ this time, in a bid to improve the United States’ cybersecurity infrastructure and to protect Federal Government networks, in increasingly turbulent and troubled times.
According to the White House briefing, the Executive Order on Improving the Nation’s Cybersecurity will make a significant contribution to the modernisation of US cybersecurity defences by protecting Federal networks, improving information-sharing between government agencies and the private sector on cyber issues, and it should strengthen the nation’s ability to respond to incidents when they occur.
It will be the first of many ambitious steps by the Administration to bring US cybersecurity into the present, but the might of the leading superpower isn’t necessarily enough; the initiatives being laid out will be dependent on collaboration between government and private bodies. The recent Colonial Pipeline incident emphasises that point ─ Federal action alone is not enough to protect every aspect of the country’s cybersphere.
With that in mind, the White House states that “much of our critical domestic infrastructure is owned and operated by the private sector, and those private sector companies make their own determination regarding cybersecurity investments. We encourage private sector companies to follow the Federal Government’s lead and take ambitious measures to augment and align cybersecurity investments with the goal of minimising future incidents.”
Key Factors of the Executive Order
Specifically, the Executive Order that President Biden signed last Thursday will:
- Remove barriers to threat information sharing between government and the private sector.
- Modernise and implement stronger cybersecurity standards in the Federal Government.
- Improve software supply chain security.
- Establish a Cybersecurity Safety Review Board.
- Create a standard playbook for responding to cyber incidents.
- Improve detection of cybersecurity incidents on Federal Government networks.
- Improve investigative and remediation capabilities.
Fortunately, Supply Chain Digital has been given some stellar insight on the cybersecurity Executive Order, from three industry-leading
Padraic O’Reilly, Co-Founder & Chief Product Officer, CyberSaint Security
Information sharing within the cybersecurity community has long been decried as something there needs to be more of. That said, it must be approached with the proper guardrails in place to ensure the protection of those sharing the information. As industries that have struggled with standardising and information sharing begin this journey, look to sectors that have successfully done it for decades─specifically, the financial services sector. By extending the guidelines seen in financial services, the disincentives for information sharing are reduced.
As the government looks to increase the communication between public and private sectors, they must work to ensure that it is a two-way street. The EO does acknowledge this need; however, historically, private sector CISOs have felt that the information sharing ends up as a one-sided relationship. I was heartened to see the urgency of the Order as well. I think the biggest challenge will be balancing the urgency with making sure that the two-way line of communication gets opened up.
The balance between the data and human problems in cyber is something we look at early and often. As a step to enhancing the posture of organisations, both public and private, the government needs to be contributing data sets such that risk management can be enhanced and performed with greater precision and knowledge. By pooling risk data across sectors, security leaders can get a complete picture which is what is severely lacking across both organisations and industries.
This executive order is a strong step toward enhancing public-private partnerships within critical infrastructure cybersecurity. The callout to the NIST Cybersecurity Framework, one of the most significant public/private sector collaborations to date, was very heartening to see and certainly serves as a model going forward.
By widening the FAR to require cyber hygiene standards across all agencies, we can begin to set some baselines. Furthermore, by learning from and integrating the DFARS and CMMC rollout within the DIB, we may begin to see the expansion of CMMC to other sectors. The critical step, though, is getting teeth behind the regulation while also making stronger cyber practices accessible. With the added language in the Order around software acquisition contracts and discussion of digital transformation of government infrastructure, it will be good to see the data collected as a result of those efforts shared with the private sector efficiently.
Joseph Cortese, CEH, Director Research & Development at A-LIGN
Although the intent of this executive order is admirable, it’s quite a laundry list. Implementing everything listed will take a very long time – especially at the pace the Federal Government moves. But here’s what really compounds the issue: yes, every step in this executive order will serve to harden the systems in question, and each of these additional frameworks will move us in a more secure direction. But it is impossible to tell if the problems we’ve been experiencing are the result of fundamentally broken systems or a failure to adopt technologies and frameworks that would have otherwise provided adequate security. Viewed through that lens, if we pile on more technology requirements that do not get adopted down the supply chain, we are no better off.
That said, there is a lot of strength in what the EO promotes. The aspect of this executive order that will have the most significant impact is the implementation of zero-trust architecture. When you look across all the controls that we use to secure technology, embodied in an ever-growing list of NIST Special Publications, it’s getting overwhelming. Zero-Trust can restructure our approach and deliver a fundamentally more secure architecture across the board.
The executive order also has its failings. One area that needs further consideration is the private sector and how they share threat information. Setting this standard will take a great deal of time and result in new bottlenecks within the private companies that conduct the threat intelligence, now subject to new requirements for feeding this information to government systems. As someone who has worked in global threat intelligence and for various agencies, the amount of information and volume of data may not be fully understood and could severely complicate the ability to execute much of this EO.
The majority of cybersecurity hacks occur due to blatant disregard for security, such as lack of two-factor authentication, egregiously simple passwords, easy-to-access software repositories, and lack of brute-force protection. What’s so upsetting to me as a cybersecurity specialist is how many of these threats can be mitigated within the private sector by increasing security awareness within organisations and by bringing attention to existing policies and procedures. It may be that greater cybersecurity awareness is the most powerful weapon we could have when it comes to the private sector.
Mike Fleck - Senior Director, Sales Engineering - Cyren
Yes, it will make a difference. Good security requires a culture of security, and culture is set at the top. This EO signals to government agencies and the tech industry that serves them that they need to prioritise security (if they aren’t already doing so). Some of the requirements have been in place for years. Most government agencies have required encryption for classified information and other sensitive data like Personally Identifiable Information (PII). There are already breach notification requirements like the ones in the HIPAA/HITECH regulation that require affected organisations to share information with their Federal Government regulators. The focus on the software supply chain is smart. We know that supply chains are and have been a common attack target.
Specifically, security standards for the software supply chain will have the largest impact. The government has been aspiring to use more Commercial Off-the-Shelf (COTS) software rather than custom-built solutions. However, it will be difficult to realise the full potential of this EO without some kind of enforcement. This EO could be similar to the process the government recently used to enforce proper security of sensitive government data stored on non-government systems. First, they published security standards (NIST 800-171) and required the defence industrial base to adhere to them. A few years later, they implemented the Cybersecurity Maturity Model Certification to enforce compliance.
Enforcement, but that will probably come in time. Also, the devil is in the implementation details. We know the government has a lot of outdated systems that can’t easily be updated to comply with modern security standards. The EO will need to include guidance for how to handle these legacy systems. We also know that government processes are far from agile. We can’t take three years to secure against the threats of the day. How will this EO incent organisations to move faster? High growth companies have long since adopted the mantra of “fail fast.” For very good reasons, the government has resisted that approach – people can die when the government fails (either quick or slowly). Finally, we can improve the security of software development processes, but software will never be 100% secure. Many, many of the large breaches have been caused by a failure to install security updates. We need to acknowledge both the manufacturer and the purchaser of the software bear responsibility to secure it.
Yes, the Software Supply Chain Security aspect should have a deep reach into the private sector. The Federal Government has the largest IT budget in the United States, so anyone selling software to government agencies should have to comply with the relevant aspects of this EO. Again, it will come down to enforcement, so “should” becomes a “must.” Security requirements without enforcement are just security recommendations.