An Industry 4.0 lens

With hardware miniaturization outpacing itself every 2 years (Moore’s law), we have seen CPUs and GPUs cramming billions of transistors onto chips that are nanometers big. The availability of highly performant computing comes at a cost, although production at scale has democratized the use of this technology in our daily lives. The explosion of cloud computing and its ease of adoption through intuitive software have powered everyday items such as smartphones and wearables, which are smaller yet more powerful than supercomputers from 5 decades ago. That said, the past decade has seen exponential growth in the availability of high-end sensors that sense their environment to process data and act (edge computing) or transmit data over purpose-built protocols, securely and reliably over longer ranges at radio frequency. This data can be relayed over 5G or the internet to data centers for further processing. With analytics and data science, we try to make sense of this data to derive further insights through AI and ML in real time. All of these advancements have brought Industry 4.0 to the fore, helping manufacturers use all these technologies to make their factories smarter, more efficient, and more automated. Furthermore, combining the operational data from production lines with enterprise data (combining IT and OT) provides elevated levels of visibility and insight for better decisions that your organizations can put into action for a competitive advantage. Essentially, Industry 4.0 underpins sustainability, as ESG requires this technology backbone.

We discussed the idea of digital entities being digital twins of real people. The same is true for equipment – we can create digital replicas of equipment, production lines, factories, manufacturing processes, or, for that matter, supply chains. With the Industrial Industry of Things (IIoT), manufacturers can simulate and improve OT to introduce new products to the market or improve their capacities by monitoring and recalibrating production lines quickly. Avoiding downtime with predictive maintenance, using AI-driven visual inspections to reduce defects and quality controls, and using robotic trained arms to accelerate critical assembly that is done manually provide the flexibility and predictability for any manufacturer to respond to market dynamics as they play out. These concepts can be applied across any industry, whether discreet or process-based manufacturing, mining, or renewables. Smart factories use simulation with digital twins and automate the entire production line to manufacture goods on demand, making any custom changes per the product design. Propelling this move forward is 3D printing technology, which makes custom manufacturing a reality and has taken the construction, food, and medical industries by storm – sustainable houses can be built to order, beverages and pizzas are made on demand, and regenerative medicine is advancing faster toward creating artificial bones and body parts for animals and humans alike. The possibilities are vast.

Speaking of regenerative medicine, these technologies also make life more sustainable. AI exponentially matures by the day and shows huge promise to contribute to human well-being. Progress in the capabilities of ML and deep learning has seen many applications, whether autonomous vehicles, predicting weather patterns, scanning medical records, diagnosing cancer earlier, or pushing the boundaries to discover novel anti-cancer drugs. Applying AI to a business application is entirely objective. It handles discreet, tangible business problems by classifying images, predicting outcomes, or extracting the key-value pairs from incoming invoices. The reliability and accuracy of an ML model depend on how clean and accurate your datasets are and whether the models are fit for purpose. As IBM CEO Arvind Krishna states, “you can’t have AI without the IA” – IA being Information Architecture. Previous chapters have stressed the importance of data architecture and creating data stores for specific functions in order to build ML models for data science. Streaming data and analytics have gained more traction with IoT connectivity and the proliferation of connected devices, with higher computing available at the edge and in the cloud. Deriving insights from big data, used as inputs to train the new ML models and improve them by the day, hones the accuracy of their outcomes.

We will now discuss some possible Industry 4.0 applications that can help with TechTronix’s challenges in launching operations and manufacturing processes for new products, and how technology solutions to ongoing challenges and futuristic plans can move Spark4Life toward its ESG goals.

Dealing with after-sales service issues

Human visual inspection is slow, often expensive, and can present quality risks. Spark4Life produces washing machines that require welded drum assemblies and drum bearings sourced from different vendors. The drum assemblies need to be inspected for a range of washing machine products and checked for defects such as cracks, inconsistent welds, and surface imperfections. Moreover, any vibration not within an acceptable limit must be observed when installed with bearings. Accepting the quality of incoming components is critical to extending the life of motors, as they spin at high RPMs. Therefore, escape rates (the proportion of defective parts that might slip through the cracks) can cause a rise in recalls and, more importantly, warranty claims, which also impacts the NPS. Traditionally, much of this inspection was done by human experts on production lines, but this isn’t optimal and can present unwanted quality risks. This makes intuitive sense; people typically don’t perform well at tasks that require focused attention for an extended period, where the probability of a notable event is very infrequent. Moreover, judgments will vary from person to person, and the time of experts is expensive too. Therefore, visual inspection AI was used on real-time images and videos of incoming deliveries from vendors. Sensors for the vibration levels were deployed for quality checks when testing assembled goods. This solution automates visual inspection tasks using image classification models and computer vision technologies that enable Spark4Life to transform quality control processes by automatically detecting product defects. The quality of the machines delivered to the market has reduced the number of service calls and warranty claims and helps increase the lifespan of appliances.

Spark4Life settles approximately 500,000 service call-outs and warranty claims for defective or damaged components. Traditionally, customers would seek repairs by calling authorized independent service providers, which were then used to send estimates and settle claims. This was prolonged and often involved disputes on whether the replacement of a particular component was warranted. This led to disagreements about charges and delays to repairs while customers were disadvantaged due to long waiting times. Spark4Life has released a new mobile app that allows customers to answer a set of questions about their problems and upload images and videos about an issue. NLP, followed by image classification, makes assessments using convolutional neural networks to provide more reliable diagnoses and request replacements with the correct spare part. There has been a dramatic improvement to first-time fix rates in these service visits, and the costs of repair (for the components) are pre-authorized in 90% of cases. Repair turnaround times have dropped from 5-7 days to 1-2 days, as less or no human intervention is required to talk to customers, coordinate with third parties, and settle balances when completing service requests. The intelligent workflow manages most of the comms, with notifications sent to the customer on the intelligently scheduled visits and timings with first-time resolution rates exceeding 90%.

The futuristic outlook of Spark4Life

As part of COP 27, Spark4Life is a key member at the table of industry majors, has pledged to net zero emissions by 2030, and has plans to introduce more sustainable practices into the production, supply chain, and circularity of its products. The aim is to work with companies to operationalize sustainability E2E by integrating and automating quality ESG data into daily workflows in a robust and auditable way. Some of the key initiatives to meet the sustainability goals that Spark4Life has defined involve launching new products and services, as elaborated here:

• Connected community sharing-based launderettes: Spark4Life has recently launched some pilot sites, first-of-their-kind launderettes-as-a-service, which allow consumers to try out the latest product models released by the company at a minimal subscription fee. The launderette sites are equipped with other older models for the community to book to use or subscribe to use monthly, with billing depending on the service tier. Subscribers can book their slots online and visit the launderette at specified times. The launderette also offers other complimentary services such as deep dry cleaning and ironing using the company’s products. With this approach, the sustainability posture of the company has improved dramatically, as the company manages all the machines and ensures the health of assets, recycling, and pre-emptive maintenance. The whole site is operated using clean energy from solar panels and has become a community spot for social hangouts, building community trust and buy-in from all customer demographics. Marketing campaigns are run for a portfolio of products from time to time to promote the company’s products and early launch offers.
• Retail stores with VR glasses/wearables: To appeal to Gen Z demographics, Spark4Life has enabled VR experiences across their retail stores in pilot regions. Every customer gets a pair of Oculus Lens 3D glasses, which allows the customer to experience VR, with a Spark4Life brand ambassador to help them with the information, queries, pricing, and other comparative studies of the product that Spark4Life offers against the competition.
• After-sales service and support: The AR app available for every customer of Spark4Life allows them to point at an appliance or product to enable a demo by the brand ambassador, who can also explain the optimal conditions to run and maintain home appliances. This has reduced the number of calls to the contact center, as many of the FAQs and service requests can be logged and answered by the digital assistant available on the mobile app. The data relating to the issue, or fault codes, are quickly relayed by the IoT sensors on the machine to mobile apps to raise a service request automatically, so the service turnaround times are minimal and more accurate when dispatching the right technician or repairing with the right components.
• Armbands for the aging population: As a part of a new health-as-a-service offering, Spark4Life plans to offer a portfolio of health monitoring wearable devices with diagnostic and advisory consultation from expert doctors, offered as a subscription. As connected devices and wearables from the company have become more sophisticated, they expose more data and services OTA, with 5G connectivity to improve patient safety, reduce costs, and improve the health of populations. This improves the patient’s experience of care (including quality and satisfaction) and reduces healthcare costs for older and younger adults. The initial pilot offers wearable devices, such as wristbands and smart watches, as part of a CSR initiative, allowing sponsored councils in the region to run the campaign for free for 1 year for all older adults. This will enable continuous, accurate data collection, subject to an individual’s agreement, for patient management and care, symptom management, medical research, clinical trials, treatment monitoring (tracking pill consumption and times), and predictive health monitoring. The data collected is also a source of information or insight that is monetized where appropriate. The added advantage of these wearables is also to track and help guide patients with dementia back to their homes and alert local authorities if they cross certain perimeters to send a pickup van to help them return to their home addresses. All the patient records and events are logged and stored in a blockchain with complete control over who accesses the patient’s data, with appropriate controls offered to patients to give permission where required through mobile apps.
• SmartBins-as-a-service: As part of a new launch of products in certain regions as part of a first pilot rollout, Spark4Life plans to launch SmartBins across all local councils as a contribution toward a greener smart city. The SmartBins are equipped with sensors to indicate how full they are. They relay the information over the LoRaWan protocol to the nearest control towers, allowing autonomous bin collector trucks to plan an optimal route daily. Moreover, citizens are incentivized to empty their food and dry recyclable waste into the correct bins and are given vouchers after a set number of compliant actions by homeowners, tracked through smart contracts. SmartBins are also located at prime spots in shopping centers, takeaways, and other community spaces, with the council leasing these SmartBins for several years, after which time the bins are planned to be recycled to make way for new, upgraded versions being made available. Spark4Life has ensured the circularity of all the components, which are sourced from sustainable resources.

Furthermore, Spark4Life is focused on establishing and implementing governance and architectural principles for their organization around processes, automation, and AI practices. It is advisable to have a Center of Excellence (CoE), with experts focusing on one single domain, which could be automation, iRPA, or CoE teams seeking out ways to optimize processes to influence KPIs, add more value, or replace them altogether after a while. Demystify concerns around automation and understand how it can accelerate work. Here, a data science CoE churns out and trains the ML models, making them available for enterprise developers to consume easily via APIs and maintaining a balance with open standard, pre-trained ML models developed by open source communities and the data scientists within the organization. The CoE is quite active in educating and training the teams on the latest trends and best practices when implementing new ideas and developing the proofs-of-concept or MVPs for selected and sponsored products.

Adopting sustainability in every aspect

Sustainability is a massive challenge for our generation and will also affect future generations. It’s centered around how data points are translated into action using technology to derive data insights – right from your installed bases (OT) at physical locations, serving customers with products and services under your new business models, to the IT you use to manage and monitor the installed bases remotely. Also, consider the supply chain that underpins the delivery of these products and services promptly to customers and how it should contribute to your sustainability goals. It’s a lot more than meeting regulatory requirements or tickbox exercises for reporting purposes; your organization should be fearless when implementing the new ideas supported by shifts technological and cultural shifts that we discussed in previous sections. Reporting and adhering to regulations will be the natural outcomes of your efforts. Moreover, these practices reflect that the brand is more ESG-aware and responsible, which has massive appeal for Gen Z and Millennials.

Moreover, embedding sustainability into your products, services, and offerings doesn’t mean completely overhauling your IT, people, and technology and rather should be planned as a slow, incremental introduction, with IT support and risk and compliance teams that will work in parallel with any transformation being undertaken and improve on it continuously. As such, it can be a natural extension of your organization’s processes. There are several cloud portfolio products from SAP for holistic management and reporting, such as SAP Sustainability Control Tower for climate action and SAP EHS for health and safety. Combine this with Ariba Network and SAP Business Network and you have E2E visibility into your ESG score and the controls to take necessary preventive actions to improve your KPIs alongside your extended partners, suppliers, and customers. When considering these products, try to answer these questions: how do you decarbonize? How do you make your net-zero transition? How do you measure your progress against the KPIs?

ESG scores

An ESG score is an analysis framework for measuring and quantifying how sustainably an organization is operating in relation to others across industries or sectors. It is a score or a metric that is used by a wide variety of stakeholders such as customers, investors, shareholders, or governing bodies to assess how responsible an organization is and how it is moving toward critical practices and taking accountability in terms of ESG. There are several frameworks, such as the Carbon Disclosure Standards Board (CDSB), the Global Reporting Initiative (GRI), the World Economic Forum (WEF), the Sustainability Accounting Standards Board (SASB), and a few others to measure how your organization fares compared to standards or other industry players. An ESG score can be used by the CxO office to understand how they are performing in terms of the ESG factors relative to their peers, or by internal LoBs so that they can compare themselves against their own annual performances or against other business units within the same organization.