Vaibhav Ghadiok designed one of the first visually-guided autonomous drones demonstrating the first instance of autonomous aerial manipulation over a decade back. At Nauto, he conceptualized, spearheaded and delivered a real-time collaborative HD mapping system for highly automated driving. He co-founded a startup developing navigation solutions using COTS radars that matched the performance of LiDAR-based navigation. He is a co-founder and VP Engineering at Hayden AI building an intelligent camera platform for smart cities.
Asif: Tell us about Hayden AI and your journey so far.
Vaibhav: Hayden AI has developed the first AI-powered data platform for smart and safe city applications such as traffic enforcement and parking management. We partner with the world’s most innovative cities to deploy our vision-based mobile solution in a city’s existing transportation fleet to collect real-time data. Our solution consists of an intelligent camera, smart cloud, HD maps, and a web portal that can be accessed by city officials.
Our mobile intelligent camera is installed in a city’s existing transportation fleet to collect data that supports the enforcement of traffic laws. In real-time, our device detects objects in the environment such as vehicles, pedestrians, lane lines, and license plates, while its Location Engine fuses data from camera, dual-frequency GNSS, wheel odometry, and IMU to track our camera’s position with centimeter accuracy. When an event is detected, it uploads a video clip to our smart cloud for further processing.
In the cloud, we build a fully annotated 3D semantic geometric map. To build the 3D geometric map, we track salient visual points in the environment. To annotate the geometric map, we segment the scene at a pixel-level and cross-register it against our 3D geometric map. This allows our camera to understand context and reason in 3D to enable it to not only detect the violation but also understand the severity and causation.
Finally, our inference engine combines these prior maps and real-time sensor data to reconstruct the scene in 3D like a game engine. We can add rules to our Reasoning Engine to fully automate the detection of any kind of traffic violation, eliminating the need for human review.
We have a fully functional prototype device for Bus Lane Enforcement running all algorithms described above. In the coming weeks, we are adding functionality to enable Crossing Guard application on our second AIoT device. We are working with our ODM partner to productionize both devices by Q4, 2020.
Asif: How is Hayden AI bringing its AI technology in a way that is different from its competitors?
Vaibhav: Compared to the fixed-function devices of our competitors, our device covers 100% of city roads, delivers 10x functionality at 10x accuracy. Our intelligent camera is capable of understanding context and reason in 3D enabling a diverse set of applications.
Asif: Please share with our viewers how Hayden AI aims to facilitate Smart city traffic operations.
Vaibhav: Everyone thinks they own the road and traffic enforcement has become overwhelmed. The solution is not just more traffic enforcement but smarter, scalable enforcement. At Hayden AI, we are creating smart city solutions purpose-built for modern traffic conditions and increased urbanization providing traffic management agencies with city-wide situational awareness. Our solution increases the throughput of current roads by enabling them to be shared with buses and bikes in a safe manner while leading to an increase in speed and ridership. We create a digital twin of the city that can provide cities with the data they need for better urban planning.
Asif: Can you shed some light on the latest employment trends related to Artificial Intelligence and IoT? Is your company hiring new talent?
Vaibhav: In the AI and IoT space, the battle for talent is fierce and especially for engineers who understand real-world requirements for successfully deploying devices in the field. The real world is complex and ambiguous and requires the building of computer vision systems that work under such conditions rather than just developing machine learning algorithms/models that achieve high test-set accuracy. Engineers that understand the interaction of device, cloud, and algorithms to deliver an AIoT solution are vital. Our company is always on the lookout for great talent.
Asif: What is the state of security and privacy with Hayden AI?
Vaibhav: Both security and privacy are built-in to the Hayden AI’s platform and we comply with all laws and regulations of any country we operate. On-edge AI processing ensures only metadata is sent to the cloud unless an event requiring video evidence is detected. And Hayden AI does not run facial recognition analysis algorithms hence protecting the privacy of pedestrians.
Our connection to the cloud is fully encrypted and we intend to use a VPN or private MVNO. Moreover, the device is tamper-proof and physical access to the device is restricted.
Asif: What other areas, apart from the city transit, do you think Hayden.ai’s technology will have an impact on?
Vaibhav: Hayden AI’s platform intends to become the nerve center of a city. In addition to collecting vision-based data that supports the enforcement of traffic laws, we also collect smart and safe city data that can be used to provide insights into hazardous driving areas, parking management, asset tracking, traffic patterns analysis, curbside management, identifying road and sidewalk hazards, and more.
Asif: Share your vision for Hayden.ai for the next five years.
Vaibhav: We expect Hayden AI’s platform to enable cities to improve traffic, reduce congestion, eliminate traffic fatalities, enable efficient transportation and achieve fair and equitable mobility for all. For instance, by introducing curbside management we enable the efficient sharing of bus lanes between buses and vehicles such as e-commerce delivery vans hence maximizing the lane throughout. Increasing the speed of buses ensures greater ridership hence ensures fair and equitable mobility. We can direct motorists to vacant parking spots hence decreasing congestion caused by motorists looking for parking. Reducing congestion helps the environment.
Asif: What are your views about MarkTechPost.com?
Vaibhav: MarkTechPost is a great source of news and trends in the ML space. It is one of the news sites I will follow.