Drives & Controls June 2022

27 www.drivesncontrols.com June 2022 AUTOMATION n mass production. They invested in developing in-house robotics capabilities by acquiring two automation companies, Grohmann and Perbix. To achieve its vision, Tesla set about automating everything in sight. It purchased more than 1,000 robots including six-axis arms from Kuka and Fanuc, and AGVs from Omron. These were applied across the process, tackling relatively routine tasks such as welding or painting, as well as novel tasks such as assembling wire harnesses. By April 2018, it was clear that things were not going to plan. The company was far behind schedule, producing an average of only 2,000 cars and haemorrhaging $100m, every week. Amid mass resignations, health and safety incidents and public ridicule, Musk dubbed the period“production hell”. The robots were not working as hoped. They struggled to hit the required throughputs and weren’t achieving quality goals. Small inefficiencies compounded and resulted in substantial delays. As a stop-gap, Tesla hired hundreds of temporary workers. Causes of failure Much of Tesla’s production hell was blamed on excessive automation but what caused such a significant deviation between expectations and reality?Why did the automation strategy fail? Tesla tried to do everything at once. It had a new product being made by a new process at a new site. On each level, Tesla was pushing the boundaries and going against orthodoxy. The first constraint that Tesla faced was the site itself. The Fremont facility had opened in 1962 and required a complete overhaul to convert it into the hyper-automated EV production machine of Musk’s dreams. Tesla had the budget for this redesign, but due to its presale promises, it did not have the time. The Model 3 was jam-packed with features that were unique in the automotive industry. And Tesla was doing this at a lower price than any of its other models. The level of innovation required many updates and changes. Due to the tight deadlines, these had to be implemented during the production ramp-up. In such a complex process, tiny changes create minor variances, which can compound and have huge impacts on upstream operations. For example, the battery cells were updated from the earlier Model S, but had a bigger capacity, so were slightly larger. An automated process had been designed to pack the cells into trays 50% faster than human operators could. Unfortunately, the change in cell size had not been considered, resulting in higher error rates. Manual operators had to take over before the systemwas redesigned. In many ways, the factory design was as innovative as the product. Tesla was relying on automation for everything from dexterous tasks such as assembling wire harnesses to heavy-duty tasks such as lifting entire vehicles. Not enough time was allowed for innovation across both the product and the processes, and for testing to ensure reliability and repeatability. The frequent changes and variability could not be accounted for, and attempts to use machine vision and other intelligent technologies were unsuccessful. Meeting the appropriate quality levels required a finely tuned production system but: n there was innovation and uncertainty across the production process; n Tesla had set ridiculous production targets requiring extreme ramp-ups; and n an unmanageable number of new, and insufficiently tested, robots made it impossible to refine each cell. Automation rethink During the Model 3 production ramp-up, Tesla’s future seemed in question. With bankruptcy a real possibility, automation seemed to be one of the main culprits. Today things are a little different. Tesla did not go out of business and successfully achieve its production goals. More than just scraping through to survival, Tesla is now thriving. The Model 3 is widely regarded as one of the best cars ever made, Tesla became the sixth company to reach a market value of a trillion dollars and, in April 2022, Musk became the richest man on earth. Patient customers and goodwill fromdeep- pocketed investors played a role in the turnaround, but in the end, self-imposed constraints and challenges led to innovations in production that have paid off a hundred-fold. Rather than lose its faith in robotics, Tesla has reshaped its approach, while doubling down on automation. The company realised that process design was much more challenging than product design. Error costs increase throughout the design process and peak once expensive fixed tooling and automation has been integrated. This equipment needs high volumes to reach profit, and high volumes demand little or no downtime. This realisation caused Tesla to change its focus – the car isn’t the only product. The factory is too and needs as much design, testing and validation, if not more. Last year, Musk explained his newmodel for production. In many ways, Tesla’s earlier challenges had come from incorrect sequencing – the production process had been designed in the wrong orer. Now Musk tries to implement a five-step process across all of his hardware companies: 1. All designs are wrong. It’s just a matter of how wrong. Everything stems from requirements and incorrect requirements have significant downstream impacts. Nail these down before moving on. 2. Delete parts of the process. Engineers tend to fall in love with technology and don’t question whether something should even exist. Ruthlessly remove features and requirements. 3. Only optimise a process once bloat has been removed. Iron out the errors and quality issues to ensure consistency and stability. 4. Only then speed things up. If the quality isn’t there, stop. 5. Automate A process needs to be stable and validated to meet a company's goals before automation can start. “I have personally made the mistake of going backwards on all five steps multiple times,”Musk admitted.“In making Tesla’s Model 3, I literally automated, accelerated, simplified and then deleted.” The last change in strategy was the realisation that automation is not a panacea. There are some benefits to“lights out” production, but full automation should not be seen as an end in itself. Instead, each aspect of the process should be assessed independently and objectively. Engineers can be drawn in by the latest gizmos and gadgets but need to acknowledge that, for now, humans are very good at certain tasks (high dexterity, high feedback, high variance). That said, the industry is evolving rapidly. Technologies that wouldn’t meet Tesla’s requirements in 2016 have advanced to the point where they are likely to – generalisable machine vision, for instance. It may have been an uphill battle but the results of Tesla’s new strategy speak for themselves. Its Gigafactory in Shanghai has 445 robots on every line and can reach peak manufacturing speeds of 45s per vehicle. This smashes Ford’s record of 53 seconds to build an F-150 truck. Tesla’s automation journey teaches us that: n Mass manufacture is much harder than low volume-production. n You need to push boundaries, but minimise risk. Unless you have Tesla’s coffers, innovating across all fronts creates risks. n Automating a flawed process makes it worse. If robotics is applied to a good process, it can be improved, but the converse is also true. n Automation brings a clear competitive advantage. When applied correctly, robotics can bring costs down, increase quality and help companies to meet demand. n * RemixRobotics isanautomationdesignagencythat buildscustomroboticsystemsforclients includingDHL, MercedesF1andtheSmallRobotCompany.Thisarticle wasoriginallypublished in itsblog, The Robot Remix .You cansubscribetotheblogfor insights intoroboticsand automation .There isa longerversionofthearticleat https://www.blog.remixrobotics.com/ teslas-automation-strategy