Plants are taken to be dumb and deaf, so much so plant communication runs counter to human common sense. But the concept of ‘talking trees’ has taken roots, and Accubits Invent, a research arm entirely funded by Thiruvananthapuram-based Accubits Technologies, has been dabbling in related research efforts for sometime.

In the pursuit of such research, Nidhin Sreekumar and his research team has stumbled upon an invention that could potentially reshape the future of Artificial General Intelligence (AGI), says Aharsh MS, Chief Marketing Officer, Accubits Technologies. AGI refers to a form of AI in which a machine can learn and think like a human. 

Critical feature for AGI 

Providing machines with a complete sensory perception is essential for AGI. It is here that, by a fortunate stroke of serendipity, Accubits Invent founded the Voltrac sensor. Its sensitivity and affordability could provide AI systems with an olfactory sense, or a ‘nose’ for future AI systems Aharsh told businessline. Initiated in 2019, the research was centred around studying the Volatile Organic Compounds (VOCs) emitted by plants. 

VOCs contain crucial information about inter-plant communications. Released by resistance-expressing plants during communication of airborne signals, they can trigger specific defensive responses in neighbouring plants of various species. 

Early experiments revealed available volatile sensors lacked in sensitivity, selectivity, and cost-effectiveness. But Voltrac has achieved a sensitivity of 12 parts per billion, exceeding the market standard of 80 parts per million. In simple terms, it’s a thousand times more sensitive while production cost is at least as many times cheaper, Aharsh explained.

Diverse applications 

Beyond plant-centric applications, the sensor is apt for diverse purposes – from monitoring perishables, air quality, and pheromones to acting as breath sensors for detecting ailments. The exceptional sensitivity makes Voltrac ideal for wearable devices, potentially pioneering early detection of serious health conditions. 

Having patented the technology, the Accubits Invent team soon realised its potential in an entirely different arena – Artificial Intelligence. While AI has significantly advanced with computer vision, audio processing, and haptic technology, its ability to replicate the human sense of smell has largely remained untapped. This is the gap that the Voltrac Sensor seeks to address. 

Aharsh spoke to businessline on the new invention.

Q

Is Accubits Invent a spin-off company of Accubits Technologies?

Accubits Invent is a research organisation completely funded by Accubits Technologies. It purveys interdisciplinary studies encompassing bioinformatics, biotechnology, nanotechnology, artificial intelligence, behavioral science, and material sciences. The research facility is established at the Bio360 Life Sciences Park in Kerala.

Q

Is Voltrac an acronym for the sensor?

“Voltrac” derives from “volatile tracking,” reflecting the sensor’s function. Volatiles are chemical compounds released by all organisms. By profiling these volatile signatures, it is possible to understand the system’s state. The name is directly inspired by the sensor’s primary function.

Q

Given its potential for applications in human and plant life-based sciences, what do you propose to do with the sensor?

The sensor has achieved remarkable sensitivity and selectivity, making it versatile across numerous domains. For example, in agriculture and the food sector, it can monitor perishable items and trace the onset of plant diseases within vast fields. This could allow farmers to address specific affected regions, eliminating the need to clear extensive farm areas or blanket-apply chemical pesticides. The sensor can also determine the optimal harvest time for crops. It’s equally effective as a breath analyser, air quality monitor, and pheromone detector, offering valuable applications in animal husbandry.

Its design can be miniaturised, enabling integration into wearable tech, such as smartwatches or fitness bands. This can potentially diagnose conditions like lung cancer, asthma, diabetes, and chronic kidney disease. This functionality mirrors the disease detection ability of certain dogs in principle, but the sensor exceeds canine sensory capabilities in both sensitivity and selectivity. Having patented this technology, we’re now poised to partner with industry leaders to cultivate production-ready applications.

Q

What are the implications for the AGI and health sectors?

For realisation of AGI, it’s imperative for machines to proficiently sense and interpret their surroundings. Computer vision imparts the ability to “see” and audio processing facilitates “hearing”, and haptic technology provides a semblance of touch. The challenge of mimicking olfactory senses has long been an elusive component in the AGI blueprint. Voltrac endows machines this capacity, and also lays the groundwork for a comprehensive foundation model for olfactory senses.

In healthcare, the sensor can provide actionable insights for maintaining wellness or circumventing health threats by analysing captured volatile profiles through our olfactory foundation model. Picture this: your smartwatch, drawing data from the sensor, advises you to avoid consuming certain foods at specific times of the day to mitigate the risk of potential health concerns.

Q

Tell us about plans, if any, to license it to corporate/private sector?

Animal and human-based use cases come with a lot of compliance challenges, we have already started working with a prominent pharmaceutical company and a biotech-focused compliance and consulting firm to navigate the compliance terrain. We have also noticed that there are plenty of relevant faster-to-market options available in sectors like agriculture, personnel security, air quality management, and AI device applications. 

While we are actively pursuing commercial applications, our research is still focused on understanding the communication between plants, the language, and on eventually even being able to talk to them in a couple of years. 

comment COMMENT NOW