It’s no secret that the COVID-19 pandemic has turned almost every industry in the world on its head and the collision repair space is no exception.
Yet this disruption, has in many ways giving the industry the chance to rethink its position and for collision repair facilities and their valued stakeholders, to look at ways to improve business efficiency, profitability and customer service.
One of the lynchpins of the business has always been the claims process. How it’s handled often determines the outcome of the repair as well as other factors such as Net Promoter Scores, business reputation, sustainability and overall success.
So, moving forward and into a post COVID-19 world, how will the claims process continue to be handled and will some of the required initiatives such as photo-based estimating and touchless transactions continue? To help answer some of these questions and more, Solera Global Marketing recently hosted a webinar entitled What’s Next: A Deeper Dive into Digital Transformation.
This event brought together a range of experts within the Solera global collision repair community, to share their views about the current situation and how machine learning technology can streamline the claims process and with it, the entire repair.
Moderated by Kyle Priest, Chief Marketing Officer at Solera, and featuring Audatex Regional Managing Directors Dave Shepherd (UK, Germany and Africa), Arnaud Agostini (France, Europe, Middle-East), Ramon Suarez (USA and Latin America) and Atul Vohra (Canada), the panel also included Elliot Roberts, UK-based Product Management Director for Audatex.
With collision repairers and insurers looking for ways to ensure speed and accuracy when it comes to claims, Dave Shepherd noted that the key to creating an effective solution is creating something that is data-driven, touchless and offers a complete end-to-end process that, not only can maximize efficiency through the entire repair but also make it easier and more straightforward for those facilitating it.
One of the key things, is having access to pre-estimate information as quickly as possible.
If we have that information within minutes of the First Notice of Loss (FNOL) it allows early, data driven decisions to be made.
— Dave Shepherd, Regional Managing Director, Audatex UK, Germany and Africa
This includes identifying the true extent of the damage, the parts and materials required to perform the repair, placing the orders to ensure they are ready as soon as the vehicle arrives at the shop and making sure the repairer and the insurer know exactly what’s involved with a particular repair on a particular vehicle and precisely how long that repair will take.
Audatex and its parent company Solera have long been recognized as innovators when it comes to estimating solutions. Considerable R&D that leverages Solera’s global footprint has resulted in the development of a digital repair platform that combines photo recognition with Artificial Intelligence (AI) and Machine Learning (ML) to deliver data driven solutions to enable fast, efficient, accurate and quality collision repairs.
Shepherd noted that by effectively mining vehicle repair data and using AI and ML to create an accurate understanding of what type of repair is required on a specific vehicle, those repairs can be triaged much faster and a number of traditional tasks automated, allowing staff in the repair facility to concentrate on working with their customers to deliver the best and most seamless experience possible.
Reduction in cycle times
Reducing cycle times is always a key factor of any collision repair operation and as Shepherd observed, COVID-19 has actually seen a reduction in keys-to-keys at many shops. “Before COVID-19, our typical cycle time was around 10 days, but we’ve seen that drop to four days,” he said.
Perhaps more importantly, Shepherd stressed that shops should take this opportunity of being able to embrace digital repair solutions and lower cycle times and make it a long-term part of their business strategy.
He noted that besides repairers, insurers can also realize significant benefits from AI and ML in the claims process to improve triage accuracy.
“It ultimately improves relationships within the collision repair network, with the customer and allows early settlements to be made based on facts and reliable data. It’s a win-win for everyone.”
Ramon Suarez stressed that the key to avoiding pitfalls during the repair process is to start by understanding the problem presented and the ability to solve it. “How we understand damage capturing solutions and how they are being utilized through the entire claims journey is key,” noted Suarez.
For machine learning technology to work effectively, it needs quality source material, therefore, providing a guided process of the type and quality of images needed to customers is critical to ensure the AI is able to augment the findings of the repair professional.
“You have to remember that you will always need expert guidance and acknowledge that every image taken can be subject to fault,” said Arnaud Agostini, indicating that smart claims and estimating technology will continue to evolve and it should be looked at as a tool to supplement and assist collision repair staff and not replace them.
Impact of total losses
Atul Vohra noted that in Canada, when it comes to claims, total losses represent 20 percent of claims but 45 percent of total severity, therefore assessing whether a vehicle is a total loss represents a huge priority for insurers.
And having a smart, efficient claims process can allow both shops and insurers to quickly determine whether vehicles can be repaired or written-off. This is critically important, since an intelligent workflow process can help huge savings to be realized. “Step back and think for a moment. One less tow is $200 saved, one day less of car rental is $100, so that’s $300 before we even start,” he said.
Not only that, but if a vehicle is allocated for repair, ensuring that it gets to the right repair facility that’s equipped with the right tools, standards, calibration and trained staff is critical.
“Because of these excessively smart vehicles we have and excessive liability issues, we don’t want unsafe repairs performed. Customer protection must be a priority.”
Vohra also noted that a persistent pain point for many collision centres is scheduling and that Audatex has developed an automated tool that aims to tackle that problem, even though scheduling is often influenced by a range of factors, including shop capacity management.
He did note however, that the COVID-19 crisis, which has allowed many shops to zero-in on their processes, will—moving forward—allow them to work with technology available through Audatex to deliver an “Amazon like experience,” for its shop customers, which in turn they can deliver to their customers—the motoring public.
Capturing the information
Capturing repair damage whether visible or hidden is critical for shops operating in a post COVID-19 environment. Yet, it’s not always as easy or straightforward as it might seem.
Elliot Roberts noted that this was something the team at Audatex were cognizant of when creating the next generation of smart appraisal tool incorporating accurate imaging. “The goal is to take a set of images from damaged vehicles anywhere in the world and do it with a high degree of accuracy and consistency,” he said.
Roberts noted that there are many factors to consider today when taking photos and assessing damage, especially when it comes to an effective automated damage assessment tool. The shape, contours and size of the vehicle, the type of damage and also other factors such as lighting conditions, reflections, dirt and water mean that taking a good quality image of the damage is a lot more difficult than it might appear.
“This was a really challenging process for us, but really the key to achieving a successful goal is access to quality data and in large volumes. For a successful AI project, especially where computer technology is at the heart. You need access to quality input data—and to this end we actually created a data lake across a multitude of countries, where Solera has access to over 1.5 billion images and corresponding repair data to over 300 million repair calculations,” said Roberts.
Roberts said that this database is updated with approximately 1.5 million images a week and when you put it all together, helps ensure the information and data remains up-to-date and relevant.
This is an important consideration for both repairers and insurers, since such as massive bank of data can help determine repair costs—a hugely important consideration in an era where vehicle complexity and severity continues to increase.
For example, Solera’s AI technology can determine that whether a particular vehicle has a steel or aluminum hood, based on the VIN data, or whether it requires a complete headlight housing replacement, or a simple bulb. “Not only is it the cost element that’s important to get right, it is the reduced cycle time of ordering the right parts as early as possible and improving the overall customer journey,” said Roberts.