Utilizing the #Multi-ArmedBandit #AI Model for Automotive #CX

Introduction

The #Multi-armed bandit model is the #AI technique which leverages large scale allocation of traffic of data, connectivity, and variations across the different screens. Named after slot machines, referred to as one-armed bandits, this technique creates an optimization strategy utilizing AI and #machinelearning.

The #Video-on-demand (VOD) and multi-screen industries have grown rapidly over the last decade creating a better user experience. The VOD market, including both mobile and cloud devices, is expanding and will continue to grow in years to come. Multi-platforms have developed with the likes of Apple, Google, Amazon, and other technology savvy players. Each have adopted middleware platform offering seamless APIs, services, and exposures. The VOD market continues to evolve and is trending toward becoming a global player with high-end devices in every industrial vertical.?

With many screens used in in vehicle design-to-production and in-vehicle-to-post-maintenance, this article focuses on the automotive industry and specifically how the multi-armed bandit model can apply to vehicles and customer experience.

MULTI-ARMED BANDIT MODEL IN THE AUTOMOTIVE INDUSTRY

The number of #Automotive sensors, actuators, controllers, and #intelligence in vehicles have increased massively in use over the last two decades. Conventional CAN, LIN, FlexRay, and MOST (Media Oriented Systems Transport) network architectures are replaced by CAN-XL and SOME-IP/Ethernet based Serviced Oriented Middleware architecture. This exponential growth of #IoV (Internet of Vehicle) components led to consolidation of #ECUs into central gateways, vehicle control units, and central computing platforms. High bandwidth, high coverage of all parts of the vehicle, and high computation needs AI like algorithms. This is leveraged for ergonomics, #comfort, #convenience, safety, and security for large firmware, middleware, application software, and over the top services and simulation. We also see a large number of multiple #screens being utilized across the entire length from steel to post maintenance phase of a vehicle’s life. These architectures are fit in a multi-armed bandit, low-safety VOD like service where it is likely to split the hardware, firmware, middleware, application, and connectivity bandwidth across the vast screens in and outside the vehicle.

USE CASES

Production support Dashboard Screens on shopfloor and supply chain

Many shopfloors use training and non-critical help dashboards during production for the homologation of the vehicles. Because the vehicle supply chain is exceptionally long, many complex parts are fulfilled in across the factories. This leaves an accumulation of data and distribution across the screens. Differing protocols, supply chain information, and part manufacturer units integrated across the multiple screens with high bandwidth display across each facility. Multi-armed bandit leverages the distribution of #supplychain data, #internetofthings data (parts and components), and #assetutilization/delivery in each dashboard during production and supply chain allocation on shopfloors.?

#CustomerExperience CX Centers at Showrooms

The key to next-generation customer experience at dealerships prior to test drives lies in multi-screen user displays. Customers can view custom-made options and features are available at different price points making the buyer-seller relationship seamless. The information is tailored given the details provided by the customer and creates an experience which removes unwanted or irrelevant information while simultaneously replacing it options the customer is most likely interested in. Large scale information and content feed based on VOD benefits from Multi-Armed Bandit for the best-in-class experience.?

#VehicleSimulators and #LicenseTrainingCenters

Vehicle simulators and license training centers use multi-screen displays facilitating the emulation of a vehicle for the driver. This utilizes high-bandwidth videos or synthetic feeds and are often resource crunched based on content delivered in VOD devices. The bandwidth, connectivity, and content give just enough for the purpose of the simulators. Because displaying all details at all times can be unnecessary and resource heavy, the operating system is able to display the most relevant information to the driver. Multi-Armed Bandit plays a crucial part in such systems used in facilities for simulations and license training centers.?

#InVehicleInfotainment and #ConnectedCockpits

Nearly all current generations of vehicles use of multi-screen displays to enhance the experience for all passengers inside. The cockpit apart from dash, center console, has things like heads-up-display, comfort screens for seat, light, door, mirror, and other adjustments. Often, too many high-definition screens including pillar-to-pillar multi-screens, and passenger infotainment tall connect to a single controller. Displays are able to identify the most important and relevant information quickly with minimal distraction. In such environments, AI techniques like Multi-Armed Bandit gives tremendous potential to fit for maximum content, VOD, and passenger experience.?

#Workshops and #RacePitstops

Racing team pitstops and workshops for road vehicles in the current generation of connected cars have enough information to be delivered from the user or the vehicle as it comes in the limp mode for service. Most pieces of information for part replacement or repair are sent to pit teams before the vehicle comes and have enough content for VOD on the dashboards for different parameters of the car. When speed and performance are top of mind, using displays efficiently becomes crucial. Unnecessary information incorrectly prioritized can hinder the driver and pit team’s ability to make strategic decisions. Multi-Armed Bandit can be utilized for enhanced information and data #contentondemand across such screens.

Visual Data

Since Visual Data on Front-end screens and VOD use high-bandwidth and high-resolution content for different use cases as previously mentioned, Multi-Armed Bandit like AI mechanisms play vital role in utilization of resources, bandwidth, and content for best user experience everywhere.

Conclusion

Multi-Armed Bandit like AI #solutions will pave way for human-centric AI in screens within and outside vehicle and such AI #algorithms are maximized for best-in-class experience within and outside vehicles for best production, vehicle and maintenance applications.

Key Takeaways

• Multi-armed bandit leverages large scale data and content both in and outside vehicle experiences demanding such AI algorithms for best performance.
• More AI based algorithms are utilized across automotive industry from supply chain to vehicles and post-maintenance. We’re able see the human-centric approaches paving way for better AI algorithms.
• In the resource and battery crunched environment, there is a need to push maximum performance with minimal workloads. Multi-Armed Bandit will see higher usage in the future for Automotive customer and VOD experiences.

About Me:

Naresh Neelakantan was part of the Core Solutions Group at a large Asian conglomerate. From last 19 years, he has worked his way in career from Project Trainee to a Senior Director in various organizations (all sizes) serving Automotive, Industrial and Embedded Electronic Industries. Being a glue guy, he facilitates solutions that exponentially transform client organizations and leverages management and technology leadership development skills to ensure the success of FPT’s delivery teams. He looks for sustainable and achievable targets, and he adds focused foresight and continuous innovation to every project. With around 19 years and 38,000 hours of experience in Research & Technology Management of Embedded Software Architectures combining Analog and Digital technologies, he has mentored and built exceptional teams, leaders, and busy beavers, at various complexities, capacities and capabilities.?