Tucked away in a bustling college town 3,000 miles from Silicon Valley is a technology start-up that boasts a considerable brain trust and equally impressive plans.
Dynadec Corp.’s outside-the-mainstream location — the company is housed near Brown University in Providence, RI — belies its mission to make optimization technology more mainstream. Founder, Chief Technology Officer, and Brown computer science professor Pascal Van Hentenryck says Dynadec’s optimization software differs from other offerings in a number of ways, but most importantly in the speed with which it parses complicated data sets to deliver decision support.
At the core of Dynadec’s solutions for vehicle routing, workforce optimization, and resource scheduling is the Comet engine. Born of Van Hentenryck’s decades of research in modeling and optimization techniques, the Comet platform uses a specialized blend of mathematical programming, constraint programming, and local search functionality to bring on-time decision support to manufacturers and others.
To produce the perfect solution to a complex problem, a traditional optimization engine might agonize over gigabytes of data for eight hours, Van Hentenryck explains. Comet’s local search capabilities instead narrow the problem set and might spit out a good enough solution in three minutes, he says.