Learn more about the benefits of mmWave radar
I flew into Detroit for the recent Automate Conference determined to put the case for simulated environments as a key accelerator for the future of autonomy. Produced by the Association for Advancing Automation (A3) this is the largest showcase of its kind in North America, staged over three days and attracting big crowds for the exhibits, keynotes, debates and presentations. So, it was satisfying that my presentation landed well – simulation is right on message and in tune with the mood of the industry.
Let me put a bit of perspective on that. Global disruption exacerbated by pandemic restrictions has left global business searching urgently for new ways to operate more effectively and efficiently. Consumer demand, supply chain turmoil and unprecedented labor shortages have understandably thrown the spotlight on autonomous systems. For many though, the scramble for adoption is being hampered by the high costs and excessive timelines associated with obtaining real-world data for autonomous mobile robot (AMR) development.
This is where synthetic data and simulated environments come into play. Using them to train machine learning models slashes the need for real-world data collection and opens the way for more ambitious and unique use cases. The Automate Conference really brought home to me that there is a growing and urgent emphasis on simulation, not just for AMRs but for automation development generally.
Many conference participants, including the likes of NVIDIA, shared their take on the simulation play – often in the shape of simulation software solutions. It was great to be able to represent the CC team with something a little different. Rather than a product with a focus on a specific application, our most recent advance in simulation is a platform that can be tailored to a wide range of use cases.
Cost-effective AV guidance
We leveraged Unity Technologies’ gaming engine to develop a modular, intelligent simulation framework for autonomous vehicle design, both for on and off highway applications. Given its use case agnostic design, it is applicable to a wide range of industries, including construction, mining, warehousing and logistics. The platform showcases a further thread of innovation from CC – a new way to unlock cost-effective AV guidance.
LiDAR is expensive, camera-based systems are susceptible to failure and interference, and GPS can suffer from scintillation (interference caused by solar flares). mmWave radar simulation represents a commercially viable addition to the technology stack, and could be a key sensor for the future of AV. You can catch up with our Innovation Briefing to explore the benefits of using radar simulation to speed up, add confidence in and cut development costs of a radar sensing array. We show how it can handle wider use cases, unlock reduced autonomous machine costs and enable greater levels of autonomy in the future.
Another theme ringing out loud and clear in Detroit was the increasing appreciation of the role that private 5G networks can play as the warehousing and logistics sector battles back from lockdown and societal restrictions. In the past, industrial wireless solutions have relied on Wi-Fi and provided an acceptable service predominantly driven by ease of deployment.
Going forward, however, the broad range of capabilities offered by customized, private 5G cellular networks has the potential to set a new global standard. It’s a solution that promises better levels of service delivered with lower latency, improved coverage and superior interference tolerance. If you haven’t already, I recommend you catch up with the CC whitepaper, Mobilizing private 5G networks, to discover more about the benefits to a large complex organization of tailoring such a private network. With companies such as John Deere already deploying them for manufacturing, I predict a wave of powerful new use cases sooner rather than later.
Automation strategy insight
My third and final big takeaway from Automate was that companies’ automation strategies need to break free from the one-size-fits-all approach – and that applies to end users like FedEx as well as tech providers like Honeywell or ABB. As I said at the outset of this blog, there’s a real post-pandemic urgency around getting on board with automation, and it’s clear that starting small and expanding incrementally is simply not delivering results fast enough. My colleague Oli Qirko has already shared some great strategic insight on this issue: Innovators must take a system thinking approach to scaling up automation.
There’s plenty of great advice in there for companies that are still quite risk averse. Many feel they are being forced to go against their DNA by the sheer pace of change and customer demand. This brings us nicely back to the theme of simulation, which can play a key role in taking cost, time and risk out of such crucial capex investments. As ever, please do drop me an email if you’d like to discuss any of the topics I’ve covered in this blog – it’ll be great to get your perspective and share in the many opinions around automation and the interface with autonomy.