Hey there, time traveller! This article was published 3/1/2014 (2208 days ago), so information in it may no longer be current.
The access to all kinds of digitized data is creating incredible changes in the way we do business.
Some of that can be uncomfortable, such as recent disclosures of random government-sanctioned peering into email and cellphone communication.
Google and other Internet utilities can target focused advertising at unique users, which is a somewhat more benign outcome of the crunching of lots of data.
Nine years ago, a small Winnipeg company called Invenia Technical Computing Corp. started taking advantage of the availability of massive amounts of free data related to the electricity markets. It started developing specialized computer algorithms to do various kinds of forecasting.
At first, those efforts were concentrated on wind-power generation, and Invenia has worked with utilities from South Africa to the U.K. and across Canada and the U.S.
But the company saw a larger niche in crunching massive volumes of data to add efficiencies to the electrical markets.
'In terms of climbing Mount Everest, we're at base camp and it's very profitable'
"We are using machine learning, a subset of artificial intelligence, which is the more general term," said Matthew Hudson, 30, Invenia's CEO and co-founder. "It is self-learning software where it takes data, references it and basically changes itself over time, evolving to do better at what you are asking it to do."
Hudson said after several years learning about the esoteric and highly complex North American electrical markets, revising its algorithms and collecting more and more data, the company is expanding its presence in the electrical markets far beyond forecasting wind-power generation.
Invenia is now using its unique domain knowledge to make arbitrage investments — buying and selling power in the day-ahead market organized around seven large independent system operators (ISOs) across the U.S. and Canada.
According to the Electric Power Supply Association, these ISOs — none of which is more than 10 years old — help to facilitate more efficient power flows and transactions by managing transmission across numerous utility areas, taking into account all the transmission and generation infrastructure in place.
After years of honing its machine-learning forecasting system — and becoming a big customer of Amazon's cloud computing services — in 2013 Invenia started making daily investments in two of those ISO markets.
To some extent, what Invenia is doing can be thought of as commodity futures trading, with the commodity being electricity and the future being the day-ahead market.
In 2014, it intends to dramatically expand that activity. To facilitate that move, it expects to be able to start raising $5 million through the province's equity tax-credit program to give it the capital collateral necessary to expand its trading activity to more ISOs.
To date, it has funded the development of its complex system internally with cash injections from all 18 of its employees and with a modest amount of angel-investor contribution.
A couple of years ago, Hudson won a Canadian Youth Business Foundation Award, and some local business watchers have pegged Invenia for big things. The complexity of what it's doing has probably kept the buzz limited to high levels.
Hudson said the company is now starting to experience dramatic growth in profits, and along with those profits, Invenia is determined to deliver a social utility as it grows.
The social utility comes from the Winnipeg company's unique ability to predict the day-ahead pricing ever more accurately. That helps to lessen the use of costly and excessive CO2-emitting thermal generation — mostly coal-fired in the U.S. — if the next-day market is read more accurately.
"For the month of November, we conservatively estimate that we prevented the emission of 13,000 tonnes of CO2 emission, something like taking 2,500 cars off the road for a year," Hudson said.
Its trading activity that month also generated about $300,000 in profits.
David Cormie, the division manager of power sales and operations for Manitoba Hydro, one of Invenia's first customers for its wind-forecasting service, said, "The technology that Invenia brings is very useful to generators, especially someone like Manitoba Hydro who would like to know a couple of days in advance what the day-ahead clearing prices are going to be.
"To the extent you have technology that looks at weather trends and price trends — there are a billion different variables that come together to determine that price — to the extent we know better what that day-ahead price is going to be, or even what the real-time price is going to be, allows us to manage our resources more effectively."
Along with co-founders Cozmin Ududec and David Duvenaud — the latter currently working on a PhD at Cambridge University, the former having already earned his at the University of Waterloo — Hudson and his team of brilliant 20-somethings have built a system that has unique capabilities to make increasingly accurate decision recommendations.
"We are planning for exponential growth," said Hudson, who is supremely confident but not cocky. "We already have it. We're just a little past the inflection point. A year from now, you will see the effects in terms of size and scale and profitability, which is really exciting."
In addition to the mostly young staffers, the company also boasts veteran resources in Nick Curry (Hudson's uncle), who is the former chief technology officer of both Great-West Life and MTS, and Kerry Stevenson, another former chief technology officer at Great-West Life.
More recently, Michael Legary, 34, the founder and chairman of Securis — a leading Internet security firm — has joined Invenia to take charge of managing the company's growth and laying the groundwork for expansion into other markets.
Legary said it is a rare opportunity to get involved in a company he believes has an opportunity to maintain very rapid growth for some time.
"My job is to figure out how we take the great tech company that it is today and put the building blocks around it for growth and scalability," he said.
Hudson and Legary said the company figures the most logical new markets it would enter would be North American gas markets, the Asian and European electricity markets and eventually some kind of play into consumer markets.
"In terms of climbing Mount Everest, we're at base camp and it's very profitable," Hudson said. "In the short term, we believe electricity markets outside North America — Europe and Asia — are similar markets with very specific, unique challenges. There are lots of parallels in the natural gas market and we will be looking at the gas market in 2014."
Martin Cash Reporter
Martin Cash has been writing a column and business news at the Free Press since 1989. Over those years he’s written through a number of business cycles and the rise and fall (and rise) in fortunes of many local businesses.
Matt Hudson on Invenia's machine-learning system and the complex electricity markets:
-- On the data Invenia collects: "The electricity grid has all these meters and all these monitoring sensors on transmission lines. It has data coming off the generating stations. There are all these points of data from the network that is the power grid."
-- On who it gets the data from: "We work with a client's SCADA (supervisory control and data-acquisition system). It is a data-based system that sucks in ridiculous amounts of data. When you look at weather, Environment Canada has weather stations everywhere. They build atmospheric models and generate simulations and they make all that available. The NOAA (National Oceanic and Atmospheric Administration) in the U.S. does the same."
-- On how management of the electricity grid has changed: "It used to be if utilities wanted to move power from one place to the other, they would have to look at who was in between A and B. They would have to make deals with all the players. For instance, if Manitoba Hydro had a great opportunity to sell power to New York, they would have to call a dozen different utilities to say we have to use their power lines and then sign contracts."
-- On how it's managed now: "Now, the independent system operators have taken over administration of the transmission capabilities so you can say move from A to B through your system and they take care of all the scheduling and planning."
-- On the day-ahead pricing market: "We are doing arbitrage, using our system and all the information available in order to understand the pricing at various points in the market that we expect to happen. We look at the difference between the current price and what the day-ahead price is expected to be and then influence the day-ahead price to reflect that. As we go through that process, we generate value by contracting that difference."