From RPA to Cobots: Automation in Insurance and Collision Repair
The potential impact of automation on the Property and Casualty and collision repair industries has become an increasingly hot topic—presenting challenges and opportunities for claims and insurance professionals. The following is a roundup of three must-know types of automation that have the potential to transform the entire ecosystem, as well as one potential pitfall to be aware of as the world becomes more algorithm dependent.
1. Robotic Process Automation
In its simplest form, robotic process automation (RPA) is the automation of repetitive tasks that humans would normally do. RPA can automate both front and back office processes, including tasks that happen all the time in the Property and Casualty and collision repair industries: filling out forms, transferring data from one system to another and taking notes. The general idea is that if a task takes a person two minutes to do, and that person typically does that task 30 times a day, he or she would have an hour each day to put to better use.
Companies will experience savings of between $5 trillion and $7 trillion by 2025.
To fully appreciate the potential impact of RPA, it is helpful to consider the data. While the industry itself is expected to be worth between $3 billion and $5 billion in the next two or three years, the potential saving far exceeds that. Researchers at Hadoop estimate that companies will experience savings of between $5 trillion and $7 trillion by 2025. Beyond that, RPA has more to offer in terms of efficiency, accuracy and freeing knowledge workers from mundane tasks in order to focus on more complex ones.
2. Intelligent Automation
Things get even more interesting when you take robotic process automation and add intelligence to it. In May, Google announced Google Duplex, an extension of Google Assistant that can make phone calls, appointments and reservations for you. The artificially intelligent personal assistant is remarkable both for what it can do and for how human it sounds. The technology feels futuristic, but it’s not far off. Google plans to begin rolling out Duplex to Pixel phones later this year.
Advancements in consumer-facing technology drive changing expectations in the enterprise.
Historically, advancements in consumer-facing technology have driven changing expectations in the enterprise. Given that, can digital personal assistants in the workplace be far off? One company, NICE, has already developed a solution called NEVA that they describe as the world’s first virtual attendant for employees. NEVA can run lengthy offline processes in the background while its human colleague provides personal interaction.
3. Cobots: Collaborative Robots
Far from what science fiction movies would have us believe, robots are not out to destroy humanity. Instead, a new breed of collaborative robots, called cobots, are already hard at work doing difficult and sometimes dangerous tasks that humans either do not want to or cannot safely do. Case in point: burger chain Caliburger’s was having trouble retaining cooks—they literally couldn’t take the heat—so they hired Flippy, an artificial intelligence-enabled cobot to do the job.
Cobots, are hard at work doing difficult tasks that humans cannot safely do.
Cobots can do much more than flip burgers: automotive manufacturers like GM use them to handle some of the physically demanding work in building cars. Companies like Amazon, FedEx and Zappos are also investing in cobots. The end game is not to replace humans—it’s to free them to do things that require greater skill, dexterity or sensitive human interaction. Take Tug, a cobot designed to deliver drugs and other supplies in hospitals. With Tug available to run errands, medical staff have more time to focus on patient interaction.
4. The Caveat: Algorithmic Bias
Algorithmic bias is the notion that because artificial intelligence is typically trained on datasets provided to it by humans, it is susceptible to developing biases based on those datasets. As the world becomes increasingly automated and algorithm dependent, understanding bias, managing it and preventing it from heading in false or dangerous directions is going to play an important role in any company using algorithms in decision making.
Managing bias plays an important role in any company using algorithms in decision making.
Case in point: MIT developed an AI entity they named Norman after cinematic history’s most well-known psychopaths, Norman Bates. They trained Norman on images from the darkest corners of the Internet, and gave him a Rorschach Test. Where normal AI interpreted and inkblot as a wedding cake, Norman saw a man killed by a speeding car. While the example is gruesome, it illustrates how training datasets can result in algorithmic bias, and thus require oversite.
The New Frontier: Collaborating with Digital Colleagues
Automation has the potential to bring big gains to the Property and Casualty and collision repair industries. It can free workers from repetitive and even dangerous tasks, enabling them to focus on higher value or personal interactions. As it does, one thing is certain: workers must learn to collaborate more closely with their new digital colleagues, whether they come in software or hardware form.