Transform New Product Introduction (NPI) with AWS to reduce cycle times and increase innovation throughput

Introduction

Welcome to our first Hi-Tech / Electronics blog post! This post begins a series of industry-focused blogs that present an AWS perspective and customer stories on how critical enterprise functions are being transformed using AWS. We are a globally distributed industry team, dedicated to helping customers around the world reinvent themselves, to accelerate their transformations, and to help drive the industry forward.

New Product Introduction (NPI) is a critical function for any customer, but especially for customers in the highly competitive and innovative Hi-Tech / Electronics industry. This blog draws on some of the successes and best practices from Amazon Devices, which runs its NPI workloads on AWS.

NPI is a top initiative for Hi-Tech / Electronics companies

According to the KPMG 2019 U.S. CEO Outlook survey, 68% of U.S. CEOs believe that it is a do-or-die for business; that being too slow leads to irrelevance.

It is easy to see why – NPI is closely tied to a company’s current and future revenue. Most companies track “Percentage of revenue from new products” very closely. According to news article in 9to5Mac dated May 2020, the latest iPhone 11 model shipments exceeded prior iPhone model shipments in 2020.

This interest is being driven by multiple factors, including a growing global middle class, with almost 400,000 people joining the middle class and every day though to 2030 (according to data from the Brookings Institute).

Challenges that impede NPI

The Hi-Tech / Electronics industry is highly competitive. Product lifecycles are getting increasingly shorter – for example some smartphones from OnePlus are now on a 6 month launch cycle.

Products are getting increasingly complex, as they strive to get smarter and more connected. Just in the past few years, we have seen the launch of entirely new categories of products like Amazon Echo speakers that do away with the need for a complex chain of audio equipment or complicated controls. Products like the Apple Watch Series 5 is capable of generating an ECG similar to a single-lead electrocardiogram. As a result, companies need an expanding number of collaboration partners.

Supply disruptions are impacting sales because they aren’t generally factored into planning / design. According to a June 2020 Bain article titled “The New Normal is a myth; the future won’t be normal at all”, a study based on more than 800 disruptions reported an average 7% decrease in sales and an 11% increase in costs, with long recovery periods ranging from months to more than two years. DfSC, or Design for Supply Chain has become an imperative.

Many hi-tech companies are seeking to make their products more environmentally friendly. According to Nielsen, 73% of millennials are willing to spend more on sustainable goods.

According to a recent survey by Deloitte and SEMI, 82% of executives reported a shortage of qualified technical candidates. Finding highly skilled engineering talent has been a challenge for some time now. To succeed, our customers are seeking to do more, with less.

Current NPI operations have not kept up. As a result, many executives are unhappy with the status quo. According to a 2019 survey by Winshuttle, over 40% seek to cut NPI cycle times by 50% or more. 60% still use email and spreadsheets to collect NPI / launch data. 37% cite lack of visibility as a challenge. NPI teams often use multiple tools and work in siloed environments, with poor end-to-end process visibility and linear collaboration models. This leads to an overall fragmented experience for innovators. A lack of collaborative tools, information visibility, fragmented tools, and overall lack of agility underpin many of these challenges.

These factors are resulting in long NPI cycle times, lower NPI throughput, low NPI success rates, and delayed revenues from new products.

How Amazon Devices transformed NPI with AWS

Amazon Devices story: Amazon itself is a Hi-Tech / Electronics company, through the Amazon Devices org. The Amazon Devices organization works on high-profile consumer electronics products and services, including multiple pioneering and best-selling products like Kindle family, Alexa, Fire tablets, Fire TV, Dash, Echo, and Alexa Wearables.

Founded in 2004, Lab126 is responsible for designing and manufacturing these devices, along with Alexa wearables such as echo earbud, Echo frame and other accessories. Over 5 years this team has transformed NPI and has been awarded at industry recognized events.

In 2007, Amazon launched one new device, the Kindle – completely on-premises, on a single site. In 2014, Amazon launched 3 new products, with roughly 50% on the cloud (Mechanical & Thermal teams). In 2020, Amazon launched over 11 new products, with teams across 25 sites, 100% on the cloud (including Mechanical, Thermal, Wireless, Audio).

Amazon Devices engineers launched over tens of thousands of simulations, covering CFD (Computational Fluid Dynamics), EDA (Electronic Design Automation), FEA (Finite Element Analysis), Wireless / RF / Antenna/ Signal Integrity, Electro Magnetics, Audio / Acoustic, Mathematical workloads.

Benefits: Increased NPI throughput 4x, reduced cycle times, cost savings, and improved agility (peak processing needs completed in 44 hours what would have otherwise taken 41 days of compute time).

Further, Amazon Devices has released its agile environment for product analysis and design to open source, so that organizations can create their own award winning designs.

https://aws.amazon.com/solutions/implementations/scale-out-computing-on-aws/

Additional details on Amazon Devices Case Study are available here:

https://aws.amazon.com/solutions/case-studies/amazon-lab126-case-study/

Additional NPI success stories

Autodesk software allows people to test ideas virtually – before you bring it into the physical world, such as simulating stress and strain, fluids, or even light. Simulations can take hours or days. Autodesk leverages AWS to achieve three levels of simulation capabilities:

Level 1 – Brute force simulations for one variable. For example, car seat design optimization, which requires a trade-off between safety and weight. Each simulation used to start with a hypothesis (a guess), requiring 3 hours for simulation. This resulted in engineers settling for an acceptable solution rather than determine the optimal solution. With AWS, Autodesk can simulate hundreds or thousands of design variations in parallel to arrive at the optimal design.

Level 2 – Generative design for multi-variable problems with a human-provided starting point.

Lightning Motorcycles an Autodesk customer, makes one of world’s fastest production motorcycles. Lightning wanted to make the “Swing Arm”, a heavy part lighter. This part is subject to tremendous force and its weight impeded performance. This is a complex problem with many variables. Autodesk leveraged AWS for generative design, with a human-provided initial design as a starting point, and arrived at a new design that was lighter and stronger.

Airbus worked with Autodesk to design Bulkheads of the latest A320. This required optimization of safety vs cost. Using generative design, and starting with an algorithm inspired by the growth pattern of an organism, Autodesk software powered by AWS generated 10,000 solutions. Each met the requirements, but with different trade-offs. The net result was that in an industry where 5% weight reduction is considered huge, the new bulkhead design was 45% lighter, and stronger than the part it replaces. If extended across the fleet, this translates to fuel savings equivalent to removing 96,000 vehicles from the road.

Level 3 – Humans only provide constraints, not a design. Autodesk customers provide inputs such as where the part attaches to the body, what forces it is subject to, what materials are feasible to use in manufacturing. The human imparts our values such as what types and levels of tradeoffs are desired – safety vs cost, weight vs environmental impact. In Level 3, AWS is not just converging – it is thinking divergently, exploring the possible design space to provide a recommended, optimized design. Once a design is chosen, AWS can help optimize manufacturing.

Watch Autodesk’s presentation at AWS re:Invent

America Makes, a leader in additive manufacturing (AM)/3D printing (3DP) technology and workforce development, leveraged Siemens PLM on AWS to realize $30M in savings, vs building an on-premises solution. Here’s what Joe Veranese, Business Systems Director at NCDMM (which manages and operates America Makes) had to say: “Had I tried, on premises, to do what we’re doing with Teamcenter on AWS, I would have needed at least 10 additional people on my team. Using Teamcenter on AWS has enabled us to keep our internal team small and focused on new value-added projects”.

Formula 1 accelerated the design of the 2021 Challenger, a new generation of race car that promises to change the Formula 1 experience for fans. By leveraging Computational Fluid Dynamics (CFD) on AWS, Formula 1 reduced the time for a 2-car simulation down to 8 hours, from 14 days for a 1-car simulation. It also tackled the problem of downforce loss for a vehicle trailing 1 second – down from a 30% loss to 5%; and for a vehicle trailing ½ – second down from a 40% loss to 7%.

Watch Formula 1’s presentation at AWS re:Invent:

AWS perspective on transforming NPI

Drawing on the successes and best practices from Amazon Devices and other AWS customers, we find it valuable to view NPI through a wide aperture – including everything from the identification of a need-gap or a new idea, to the first satisfied customer.

The start of the process begins with the identification of a ‘need-gap’ in the market, with a new idea, or with the start of an iterative NPI cycle of an existing product. Here, many customers leverage solutions like AWS data lakes, AI/ML for insight generation, and Amazon Forecasting to estimate demand.

This leads to concept definition. Here, many customers are curious about how Amazon and AWS innovate, where we typically kick off NPI with our Working Backwards method.

For the Design Simulate and Test, many customers are converging on the cloud as the best place for NPI workloads – even for those they deemed nearly impossible or challenging even a couple of years ago. There are many reasons for this change in thinking. Leading ISVs – including CAD, CAE, and PLM, are available on AWS. Customers are performing advanced simulations at scale that are simply not feasible, or cost prohibitive to do on premises. Customers are realizing new possibilities through generative design – where high performance computers leverage AI/ML to recommend optimal designs, with limited input from humans. They’re leveraging services like AWS SageMaker for AI/ML, Amazon Kendra for knowledge engineering, various ISV and bespoke tools for running their Computational Fluid Dynamics, EDA, FEA, Wireless / RF / Antenna / Signal Integrity, Electro Magnetics, Audio / Acoustic, and Mathematical workloads.

NPI does not end with design. To create that first satisfied customer (and beyond), NPI needs to incorporate MES and ERP as well. Leading MES ISVs and ERP vendors are all available on AWS. AWS makes it easier to train machines and robot, to help them conform to the real world (rather than vice versa), and optimize manufacturing and order management processes. As a result, AWS is helping manufacturers scale and deliver quicker than ever before.

Looking beyond the NPI process steps, customers are re-thinking how they collaborate. They’re looking to speed up collaboration by breaking free of linear models, and moving towards Dev-Ops inspired models, or neuromorphic models that mimic the human brain. AWS Secure Collaboration Chambers are already being leveraged by Amazon’s own Annapurna Labs and many other customers for Electronic Design Automation. AWS is continually innovating around new collaboration models in conjunction with our customers.

AWS customers are making their end-customers an integral part of NPI through continual feedback. And to close the loop, customers are measuring customer satisfaction and continually improving their products using AWS.

AWS is making all of this possible by building secure tools that enable end to end visibility and easy data sharing across multiple partners.

Summary

The Cloud isn’t just about lifting and shifting NPI workloads to reduce expenses. AWS creates opportunities to reimagine the customer experience, and to create, deliver and manage products that will create an experience in ways that are simply not possible otherwise.

Getting started with NPI on AWS

Ask about our collaborative AWS NPI Workshop, which helps us establish NPI goals, identify bottlenecks in NPI, and identify use cases to maximize impact.

Contact: vijaywun@amazon.com