MDAP, or Mobile Data Acquisition Platform, is a new addition to your suite of innovative tools. What can you tell us about it?
Mitra Shah: We developed MDAP to form part of ADEP – our Automated Driving Engineering Platform. MDAP is a customizable vehicle-based sensing system comprising a camera, radar, lidar and GPS. Imagine a small rooftop box that can be mounted on top of a car, or, in fact, a bunch of other locations. We originally created MDAP mostly for developing advanced driver-assistance systems (ADAS), but soon found that it can be used in a wide variety of situations. As well as on-road vehicles, we can also use it to gather data for offroad vehicles in industries like mining and construction. MDAP can also be installed at traffic intersections, and we can even use it in different workplace scenarios like factories to monitor employee movements.
MDAP, through its innovative use of computer vision, is a key enabler to provide and extend our services across the Intelligent Transportation Systems sector enabling technology development and evaluation via computer vision. By leveraging this technology, we are not only able to push the boundaries of automated driving development but also significantly enhance Traffic Monitoring and Analysis, Worker Safety and Tracking. These areas are deeply dependent on precise and comprehensive Computer Vision Data for development, operations, and detailed analysis.
Can you give us more details on how it helps in those different situations?
Mitra Shah: For the automotive sector, we use MDAP to help develop and evaluate ADAS such as adaptive cruise control, emergency braking or blind-spot warning. By mounting it on top of a vehicle, MDAP can help give incredibly accurate data on things like braking speeds, distances, reaction times and so on.
When we use MDAP to monitor traffic intersections or specific stretches of highway, it can capture real-time data on things like traffic flows and pedestrian movements. The Data that Is collected with insights can then be used for V2X, or “vehicle-to-everything” communication, so connected vehicles can ‘talk’ to each other, which can help make driving safer and more efficient Traffic flow. V2X is a vital part of semi-autonomous and autonomous vehicles. This traffic data can also be hugely beneficial for city authorities for planning and Analysis. Accenture’s platform, for instance, could ingest data from MDAP and then deliver insights into how to make an intersection safer, for instance.
And MDAP can also help make working environments run more safely and efficiently. Install it at a factory, warehouse or chemical plant, for example, and it can monitor how employees and robots move and interact with machinery. This data can then be used to adapt safety procedures or adjust workflows for greater efficiency.
MDAP is a great way to capture “ground truth” data. Can you explain what that means?
Samer Labban: Ground truth data is data that Is built on the raw input coming from the highly accurate sensors onboard MDAP. This raw data can be curated to create datasets to both train and validate machine learning systems in advanced driving technology. Ground truth data is also used to calculate metrics for Key Performance Indicators (KPIs) of ADAS features. MDAP’s sensor suite is too big and expensive to use in standard production vehicles; however, the rich context of the combined data from these sensors can be used to evaluate and improve the safety and performance of these vehicles' ADAS features. Companies need to know how to leverage data driven approaches to develop and evaluate autonomous features and accurately perceive the environment – which is where ground truth data shines.
What was the inspiration for developing MDAP?
Samer Labban: We entered the Autonomous Driving Sector to make our contribution towards safer and efficient driving, and MDAP is a representation of those goals. Our first milestones were centered around “sensor fusion.” This is the process of combining data from various sources to produce more reliable results than if these sensors were to be used individually. We soon realized that our design translated well across multiple sectors and could make an impact at a greater scale. It’s important to stress that this is about more than “just” hardware. MDAP feeds into our ADEP ecosystem, which is used to develop and evaluate autonomous vehicles and associated technologies like V2X communications and infrastructure traffic monitoring.
How do you expect MDAP to evolve in the future?
Mitra Shah: We’re working on a new tool that combines the data captured by MDAP with simulation data for CAV evaluation and also traffic analysis. Long term, it’s not so much about the technology used to collect data, but how we leverage that data. So, no matter what comes up in terms of sensing capabilities, we will always be able to leverage the data with the tools we have in the background.
Thanks for the interview!
