Top 5 Enterprise AI Costs in Manufacturing
For most manufacturers, the goal of implementing a data-driven AI strategy is to harness large, heterogeneous datasets in order to improve manufacturing processes. For some, the end goal may be optimising manufacturing scheduling and asset utilisation in order to understand and predict scheduling issues and root causes.
For others, AI may be the solution to quality management. By using data to reduce the cost of quality checks, failures and reworks, you can optimise settings and production parameters. For others still, AI is intended to drive real-time reporting and analysis, as it can unveil patterns in manufacturing processes and supply chains, allowing manufacturers to predict and act in real time.
As with most manufacturing use cases, the tenth or even twentieth applied AI business use case will generally still have a positive impact on the balance sheet, but what most are finding is that the marginal cost of supplemental use cases does not decrease, while the marginal value of the use cases does in fact decrease. Most manufacturers in this position will quickly reach a point where the marginal profit of new AI use cases enters the negative.
AI needs to drive efficiency for manufacturers and ideally, in today’s challenging economy, it needs to be a revenue centre, not a cost centre. It’s not enough for organisations to leverage AI at any price. If AI is to be truly sustainable, we need to deal with some of the less tangible costs that add up over time and, therefore, hinder manufacturers from scaling and profiting with AI.
The 5 Costs Hindering Enterprise AI and What to Do About Them
1. Data Cleaning and Preparation: Most data teams will agree that data cleaning and wrangling is the most difficult or time-consuming part of data processes at their organisations. The real problem with data prep is that organisations are doing it separately for every single use case and project.
Instead of having repeated (and therefore costly) work across people, teams, and the wider organisation, manufacturers need to prioritise data prep efficiency. This means putting systems in place that allow data to be found, cleaned, and prepared once, then used a maximum number of times across different use cases.
2. Operationalisation and Preparation: It’s not uncommon in a manufacturing data science lab for the first version of any machine learning model to take six months from release to production. The problem here is also one of scale: think about the lost revenue for the amount of time a given machine learning model is not in production and able to benefit the business. When an organisation wants to scale and implement hundreds of models, the cost is debilitating.
Manufacturers need to invest in establishing consistent processes to manage the packaging of code, release, and operationalisation. By incorporating reuse from design to production, manufacturers will be able to scale without the need to recode models and pipelines from scratch in order to operationalize.
3. Data Scientist Cost and Retention: There are also people costs to deal with while data scientists spend time in the lab prepping prototypes to work in the real world. Data scientists, by nature, want to be executing on models. Many organisations find they may lose talented data scientists along the way because they couldn’t streamline operationalisation and production.
Don’t make data scientists do things twice. Reducing costs associated with the inefficient use of data scientist time is usually a matter of introducing proper tooling along with the right operational processes to improve efficiency. Companies need to provide the resources for data scientists to reuse work and lessons learned from past projects.
4. Model Maintenance: Even once models make it into production, many manufacturers encounter difficulties in maintaining them, as data is always changing.
The Solution: The more use cases a manufacturer takes on, maintenance becomes even more challenging — which drives up the costs even further. MLOps has emerged as a way of controlling the cost of maintenance, shifting from a one-off task handled by a different person — usually the original data scientist who worked on the project — for each model into a systematised, centralised task.
Part of maintenance is also ensuring easy reuse across organisations, which means that people can easily access information and consume things done by others, including seeing data transformation and models.
5. Complex Technological Stacks: Model maintenance itself is complex and can be costly if not properly addressed, but this is often also the case with the infrastructure and AI technologies that various parts of organisations are using, especially if they are geographically dispersed.
The Solution: Organisations need to be able to stitch together their larger technology pictures, as without this ability, they won’t be able to reuse and share knowledge across teams, which usually leads to additional costs.
Shifting to AI as a Revenue Centre
In the coming years, the ability of manufacturers to pivot their activities around Enterprise AI will fundamentally determine their fate. Those able to efficiently leverage data science and machine learning techniques to improve business operations and processes will find new business opportunities and get ahead of the competition, while those unable to shift will fall behind, perhaps swept away with the tide of rising costs and diminishing revenue. To achieve this, we need to move AI from a cost to a revenue centre.
If manufacturers can address the costs outlined above and decrease both marginal costs and incremental maintenance costs, they will be in a strong position to apply reuse models and will find scaling a data-driven AI approach much easier in the future.
5 Minutes With PwC on AI and Big Data in Manufacturing
Please could you define what artificial intelligence is, and what Big Data is?
AI is the ability of a machine to perceive its environment and perform tasks that normally require human intelligence, and it’s a whole field of different technologies, techniques and applications.
Big data is a set of tools and capabilities for working with, for processing, extremely large sets of data.
How does AI and Big Data work together?
Big data is just one of the enablers of AI, though as we see increasing volumes of data, it’s one of the most important
How can this be applied to a manufacturing setting?
Broadly speaking, there are many benefits of AI, and the use of data, which include reducing costs, minimising human error, and increasing productivity and efficiency. The important thing to consider is any setting - for the use of any technology - is what is the problem you are trying to solve? Be it merely automating repetitive tasks or to reinventing the nature of work in factories by having humans and machines collaborate in order to make better and faster decisions.
Why should manufacturers use AI and Big Data when adopting smart manufacturing capabilities, what is the value for manufacturers?
One view is, again, the economic benefits of AI, which come in manufacturing as a result of:
1. Productivity gains from automating processes and augmenting the work of existing labour forces with various applications of AI technologies.
2. Increased consumer demand due to the increased ability to personalise and tailor manufactured products, along with higher-quality digital and AI-enhanced products and services.
Manufacturing (and construction industries) are by nature capital intensive, and in our 2018 report, “The potential impact of AI in the Middle East,” we estimated that the adoption of AI applications could increase the sectors’ contribution to GDP gains by more than 12.4% by 2030.
How can AI and Big Data help manufacturers to evolve in the Industry 4.0 revolution? What about those already looking at Industry 5.0?
It’s really about the investment you make now, in order to futureproof your business.
We typically see two broad strategies or approaches to the adoption of AI. There are things that we can do immediately, without any recourse to Big Data - which is to adopt technologies we describe as Sensing, those involving computer vision, for example. There are plenty of use cases where these can be used immediately in manufacturing, such as for automatic fault detection. However, there is a longer term play which requires investing in data - getting the right collection mechanisms in place, storage, data governance, Big Data capabilities etc - in order to develop increasingly valuable machine learning driven AI use cases. This is absolutely necessary for long term adoption success.
What is the best strategy for organisations looking to realise the value of AI and Big Data in manufacturing?
AI and Big Data are only one part of a successful smart factory. The organisations that lead on AI adoption are those who have already made the most progress in digitising core business processes. In order get ahead in using AI solutions at scale, there are a number of technology investments and organisational decisions to be made, including:
1. Digitising processes ultimately leads to improved ability to generate data, and in the manufacturing setting - with many 100s of sensors generating 1000s of measurements in real time, the result is Big Data. Data is key to building AI so reliable and accurate data acquisition, management and governance are key. The production line and factories play a critical and direct role in the data-acquisition process.
2. AI strategy, both long and short term, begins with the use cases, the business applications. Manufacturers need to ask where they want to use AI and gather these use cases together and prioritising projects based on a balance of expected impact and complexity of implementation.
Of course, in addition to technology and business processes, people are at the heart of any successful technology adoption. AI teams need to be composed not only of data scientists, also data engineers and solution architects to enable their work, data stewards to ensure accuracy, and increasingly so call “Analytics/AI translators” who are able to communicate with business leaders and technology experts. Culture is also key, and manufacturers need to enable a data and AI-driven culture, building trust in data and algorithms by educating their workforce about AI and its capabilities, how best to extract value. It’s not just the positive of course, but also the risks and limitations, as these when encountered without expectations having been set, can significantly impact willingness to invest.
What are the challenges when it comes to adopting AI and Big Data in manufacturing?
has shown that one of the major challenges to implementing AI is uncertainty around return on investment (ROI). As I said, there is significant investment required for a long term data and AI strategy to be successful, and expectations around the time to see tangible returns must be set realistically.
Many companies also struggle with the data side: collecting and supplying the data that an AI system needs to operate, and ensuring that it is accurate. Again, this speaks to the bigger investments required in digitisation.
Some of the main challenges for manufacturing companies with implementing AI at a scale from our research include:
- 40% → Technologies not mature
- 40% → Workforce lacks skills to implement and manage AI
- 36% → Uncertain of return on investment
- 33% → Data is not mature yet
- 32% → lack of transparency and trust
- 24% → Work councils and labour unions
- 22% → Regulatory hurdles in home & important markets
One element highlighted here, particularly around lack of trust, and labour unions, is that AI is typically misrepresented in the media as “replacing” workers, and taking jobs. Yes, there are efficiency gains to be made from automation, as there have been since the first industrial revolution. But we believe that Data and AI are at their most valuable when they are used to augment workers, enhancing their abilities and the products being manufactured.
Another challenge we’re starting to see emerge is cyberattacks increasingly targeting interconnected equipment and machinery in smart factories. PwC recently hosted a webcast, in cooperation with the National Association of Manufacturers in the US and Microsoft to discuss this.
What are the current trends in AI and Big Data in manufacturing?
- We see companies putting slightly more focus on adding AI solutions to core production processes such as the engineering, and assembly and quality testing
- Safety is of significant importance, with techniques adopted in protocol adherence capabilities (for example maintaining safe distance from specific machinery) being adopted in more every day scenarios for COVID-19 protocol adherence
- There is considerable interest in predictive maintenance for large machinery involved in manufacturing processes, and also supply-chain optimisation
What do you see happening in the AI and Big Data industry in manufacturing in the next 12-18 months?
Honestly, I think we’ll see a continuance of where we’ve already been going for the last 12- 18 months. AI and data are already being used in manufacturing but this use doesn’t get as much attention in the media as, say, healthcare, but the success stories are there, and they will continue as operations continue their digital journeys.