Key Technology Applications in Automotive Manufacturing

While the automotive industry carries a reputation as conservative and as slow to adopt innovation, that only applies to its final products. Even then, the industry explores new technologies relatively early, but because its products have long life cycles, it often takes several years for the new technologies to hit the market. When it comes to Smart Manufacturing technologies, the automotive industry has served as one of its pioneers.

The automotive industry has seen its share of publicity stunts, such as Audi’s Factory 4.0, but manufacturers have truly made strides in adopting transformative technologies in their factories. These new technologies range from Additive Manufacturing (AM) and Augmented Reality (AR) to Artificial Intelligence (AI) and Collaborative Robotics (cobots).

The use cases below and case studies only include a very small portion of the Smart Manufacturing applications that are already implemented in the automotive industry.

• Additive manufacturing for Original Equipment Manufacturers (OEMs):BMW ordered the first EOS Electro Optical Systems stereolithography 3D printer ever sold back in 1990. Now, it uses EOS printers for a folding bracket for the convertible BMW i8 Roaster soft top. The part is now 10 times stiffer and 44% lighter than the alternative plastic injection molded part. Mini, a BMW subsidiary, uses EOS printers for customized interior and exterior components. Rolls Royce Motor Cars, another BMW subsidiary, uses EOS printers for production parts. In the Volkswagen (VW) Group, Audi uses EOS printers for production toolmaking. Daimler, which has 317,000 active spare parts, now uses EOS machines to produce parts for its EvoBus on demand, reducing inventory costs. EOS estimates it has 85% of the additive market share with OEMs and about the same with Tier 1 and Tier 2 suppliers.
• AR on a minor supplier’s production line:WS System, a German automotive parts supplier, uses Vuzix M300 smart glasses with Ubimax AR software connected to sensors on two assembly lines. The line automatically stops if weight sensors below the boxes of inventory do not feel that the parts taken out of the boxes at the correct time. The glasses then alert the workers to the error. This has sped up training, increased operational efficiency, and decreased errors.
• Connecting equipment: In its Alabama plant, Honda has Omron, Rockwell, and Mitsubishi machines and a Manufacturing Execution System (MES) built in-house. They are needed to reduce custom coding to free up more of their Information Technology (IT) staff’s time. Even with OPC Unified Architecture (OPC UA), it still had too much custom code and did not significantly lessen the demands on the IT department. Only after adopting Telit’s deviceWISE did they find the simplified architecture and code-free app development they needed. Telit provided the application integration, and Honda only needed to configure the logic to its needs. The codeless development added to the protocol translation and edge intelligence provided by Telit, resulting in easily scalable apps and solutions. It also improved the speed at which Honda could update or change their production lines, increase or decrease production and reconfigure Programmable Logic Controllers (PLCs).
• Manufacturing execution for engine supplier:Dassault Systèmes’ DELMIA provides the Manufacturing Operations Management (MOM) system for the engine manufacturer Cummins. Cummins shifted from directing all of its manufacturing systems on paper to doing the same through DELMIA. As a result, Cummins saw a 90% drop in quality claims with a 25% production throughput increase. It has now deployed DELMIA in 22 plants, with two more in progress.
• Paint shop quality monitoring:Software AG has an automotive manufacturer as a client who uses its Apama product for streaming analytics and its Cumulocity Internet of Things (IoT) to train Machine Learning (ML) models for quality monitoring of the paint jobs. The ability to detect defects in the paint jobs in real time and to fix the mistakes or missed spots makes the function of metrology/Quality Assurance (QA) commensurately active (versus reactive).
• Process planning for automotive OEM:Honda might use Telit to connect its equipment, but for process planning, Honda North America uses Dassault’s DELMIA. Honda has worked with DELMIA since the early 1980s and has fully committed to digital manufacturing. It uses DELMIA to model and simulate everything on its plant floors before physically building it.
• Screwdriving cobot applications for OEMs and suppliers:Universal Robots did not think that the automotive industry would adapt to cobots, but it now supplies cobots to 90% of all OEMs. Some, such as PSA Group, have scaled deployment of Universal’s UR10 cobots for screwdriving applications on the assembly line. It sells even more cobots to automotive suppliers such as Continental and Lear. The cobots work alongside humans and have improved ergonomic work conditions and productivity on the lines. The screwdriving applications reach a Return on Investment (ROI) in an average of three to four months.
• Smart Manufacturing platform for automotive parts:HIROTEC, a Japanese auto parts manufacturer, needed to eliminate unplanned downtime. It adopted PTC’s Kepware to connect and integrate its equipment and PTC ThingWorx in its Detroit plant as an on-premise cloud platform running on HPE Edgeline systems. HIROTEC has connected its Computer Numerical Control (CNC) machines, inspection robots, force sensors, laser measurement devices, cameras, and robotic arms. It now generates automatic reports for the entire production line and predicts and prevents failures in critical systems using ML. It has plans to continue to leverage the data in extended reality apps.
• Windshield glazing:When producing windshields, the glass often differs slightly in geometric shape, even for the same model of car. Different environmental conditions can affect this, making it difficult to predict how the glass might vary. The different geometries can disrupt the operations of the robots applying adhesion at the primer station, causing production to recalibrate the robots about three times each day. By leveraging edge computing on Amazon Web Services’ (AWS’s) Greengrass, a Boston Consulting Group (BCG) Digital Ventures interdisciplinary team developed a “digital shadow” of each individual windshield. A KUKA robot cleaned the windshield before it went to the primer station. While cleaning, it can now measure the 3D shape of each windshield, create a 3D model, and then save this model as a digital shadow on an edge device. The robot applying adhesion then adapts its programming to fit the exact shape of each individual windshield. The digital shadows sync with the Enterprise Resource Planning (ERP) system and a cloud platform based in AWS whenever connectivity allows for the training of the robots’ computer vision for quality control. Any updates to the algorithms then deploy to Greengrass at the edge.

To flexibly adapt to automotive manufacturing clients’ needs and to help them scale, Smart Manufacturing vendors must:

• Meet with different stakeholders.The automotive industry carries a weight of publicity, volume, complexity, and national pride that most other industries do not. Automotive workers unions wield immense power in countries such as the United States and Germany. Governments and political leaders pay attention to their OEMs; in some cases, this means navigating health and safety regulations, especially around cobots or AR on the production line. In other cases, such as in China or France, it might mean the government directly invests in or subsidizes the industry. Smart Manufacturing vendors should consider and talk to the stakeholders in each individual company they target—especially in the automotive industry. Obviously, this includes the IT, operations, research and development (R&D), and other relevant departments within the company itself.
• Prioritize the business case. High value, high complexity products, downtime, and loss of productivity come at a much higher cost while incremental increases in productivity or quality pay off quickly due to the automotive industry’s high volume of manufacturing. If vendors cannot guarantee ROI at scale, manufacturers in this industry will not implement their solutions. This industry does act as a pioneer but only for solutions that can prove their value.
• Build—or partner with a vendor that builds—process simulation technology. Considering the cost of test vehicles and the demand for ROI, process simulation tools can prove extremely valuable in demonstrating the benefits of new technologies in production processes, identifying bottlenecks, meeting health and safety standards, and achieving first-time-right production. Most clients will still want to go through physical test runs, but simulation technology can reduce the failures of those test runs.
• Stay hardware agnostic, spend time with customers’ IT and Operational Technology (OT) professionals, andleverage existing infrastructure. Automotive manufacturers tend to have some of the strictest and least flexible IT policies on the types of devices or data that connects to their networks. Vendors should meet this standard with cross-functional teams from every new client to determine how to adapt solutions to both IT and OT needs. For software vendors, this means staying hardware agnostic and using whatever devices the client already has or already knows how to integrate.

Following this strategic guidance should help Smart Manufacturing vendors and their clients implement scalable solutions.

For more insights and perspectives on manufacturing and the industrial Internet, please check out ABI Research’s Smart Manufacturing service.