Mediaboss Marketing

Avalo's machine-learning tech speeds the development of new crops.

Climate change is a big problem that requires big solutions, but one place where it might have an unexpected impact is on your dinner plate. By making climate-resilient crops with the help of machine-learning approach, Durhams Avalo looks to keep those plates full.

The companys Chief Scientific Officer, Mariano Alvarez, will present tomorrow (March 29) at this years Venture Connect summit in RTP.

Like many great ideas, Avalo was conceived between a pair of friends over a couple of pints. Scientist-turned-entrepreneur Brendan Collins finally convinced his friend, Duke post-doc researcher Alvarez, that his research on plant genetics could do a lot more good in the real world than in the lab.

Over a series of happy hour beers, he convinced me that it would be way more fun to start a startup and do research in that setting than to continue to apply for faculty positions, Alvarez said. So in 2020, we did just that. And he was totally rightits way more fun.

Alvarez is used to tackling plant genetics in light of climate crises. In the wake of the Deepwater Horizon oil spill crisis in 2010, he completed PhD research at the University of South Florida looking at how plants reacted to the dramatic environmental changes.

This research brought the young scientist to Duke for post-grad study on understanding the relationship between plant genomes and the environment. Guided by Duke computer science professor Cynthia Rudin, Alvarez soon realized that machine learning and computational methods could solve many of the problems in identifying genes that are meaningful to plant environmental resilience.

Around the same time that the Duke duo figured out how to use machine learning as an impactful crop-development tool, Alvarez and Collins began doing market research. The need for faster crop development was urgent, they foundand with climate changes biggest effects just decades away, the time was ripe to launch their startup.

A lot of people dont realize just how long it takes to come up with a new variety of crops, Alvarez said. You go to the store, theres different types of tomatoes, theres different types of cucumbers, and you sort of imagine that theyre all just sitting around. It actually takes a long time for somebody to develop those varieties, anywhere between seven and 15 years. Its roughly a $200 million process to actually get them through trials and into farmers fields.

Collins, who is Avalos CEO as well as co-founder, brought his software-scaling skills from previous startup ventures, allowing them to translate this computational model into a marketable product.

Using its computational model, Avalo can rapidly test for genes that may produce a desired phenotypic outcome in a plant. The companys computational engine allows them to discover the genetic basis of complex traits, even from patchy data.

This not only makes the process of developing new crops much faster, it also makes it cheaper by slashing the number of years needed for research and development.

Traditionally, crop development has focused on traits that will make a process that takes 15 years and $200 million worth their while, usually aiming for genetic variations that lead to high yield or herbicide resistance. With Avalos technology, companies can focus on other traits, like ones that make a crop able to grow effectively in the new temperatures that result from climate change, or even tweaking a crop for better taste. This technology is perfect for a diverse industry with diverse needs.

One thing that was really interesting going into agriculture is just the scope of all of the things that people are looking for, and how diverse the agricultural system is, Alvarez said. And how unique growers needs are.

Avalo offers three buckets of product, Alvarez says. In the computational bucket falls Avalos work in providing their computational tools to companies who know what traits they want but dont have the technology to make it happen. In the second bucket, Avalo transfers specific traits into a plant for a customer.

In the third bucket is Avalos front-to-back operations. The company is currently working on a heat-tolerant variety of Chinese broccoli, for example, but their capabilities are not crop-limited, especially with the help of their three greenhouses spread across the Triangle.

The company is growing fast, but then again, so is the problem it looks to addressclimate change waits for no man, or machine.

We really only have about 30 more years until some of the biggest changes start to take effect, Alvarez said. If development takes 15 years, then we only have two shots, which is just not enough time to develop the varieties that we think were going to need to to adapt an entire agricultural system to a whole new climate.

See original here:
Durham's Avalo Uses Machine Learning To Let It Grow - GrepBeat